I feel this HN account had been just waiting for this article.
The points the article makes about torsional and lateral tension helped me understand better why I prefer steel over carbon. I've always described it as "feeling more in-sync" with the bike. For example, the subtle instinctive biomechanical action of putting more weight down on the outside pedal to re-align the wheels when you start to understeer just feels great, like you're "one" with the bicycle.
In bike messenger circles there is a phrase "steel is real". The idea being that when you inevitably eat shit going around a corner you might dent your frame, but the chances that a steel frame will crack or shatter is basically zero. And more importantly dings and scratches don't cause the structural integrity to degrade like happens with aluminum or carbon fiber. Your steel fork is bent? Hit it with a hammer until it's true again!
A year ago, my titanium fork broke (while driving rather slowly on good tarmac, photo here https://twitter.com/dothebart/status/1486765955254530051), and the titanium frame got a few cracks as well. My motto "the last bike I'll ever buy" collided with reality. Fortunately, I got away with just a few scratches - just a minute later, I would have been at 60+ kph downhill.
My new bike has a steel frame. I love it, but titanium was more flexible and more comfortable. And of course titanium is nerdier! :-)
One of my coaches in college raced for a pro/semi-pro team. Their first race of the season was on new/pre-production bikes straight from one of the major manufacturers. 3, I repeat, 3 steer tubes broke in that 1 race. It was a really rugged race, but needless to say the manufacturer had to scramble to fix the forks before going to production.
One guy was going over railroad tracks and his handlebars just came off. Not what you want in the middle of a cat-1/pro peloton.
I had one of the fork ends break off a steel fork. Fortunately I was going at less than a walking pace, navigating out of a store parking lot. I was 2 miles from home, and just walked the bike while planning its replacement.
It was a break that I would have noticed much sooner, had I regularly inspected it, because the other side was half broken off.
The funny thing is, I play in a band, and when I told my story that night, 3 guys in the band offered to weld it.
I'd get the drummer to put together the rig which holds the parts in place, bass player to make the crucial spot welds and then the lead guitarist to go ape shit with the rest, while the singer goes whoo yeah!
Yikes! Fork failure is the stuff of nightmares. I'm glad you're okay.
Among the participants of a local riding club in my area, I've met two people with custom titanium frame builds who told me that their frames cracked within a few years of use.
Still have nightmares about a fork that, during a mountain bike descent 30 years ago, started oscillating increasingly until it bent and locked the wheel perpendicular to my vector - and my buddy still has dreams of me acting Superman along his ride, for a few short instants...
Titanium can have problems with oils I've been told. Seeping into crystal defects and propagating. Typically I've been told they use polymer coatings to prevent the ingress. But with scratches and lubricants I think this would be much more common if it was a real thing.
I've always thought a spring bronze/brass frame would be fun. Even better modulus of elasticity, but the strength to weight is much worse.
Did the manufacturer provide an explanation? A sheer (which that looks like) should be near impossible on any properly manufactured bicycle under normal usage.
I wouldn’t view that as an indictment of a particular material.
That’s like first generation carbon failure and those went away 20 years ago. I’d blame the manufacturer specifically in that case because it’s both catastrophic & unexpected.
The “anti helmet crowd” wouldn’t have any issue reading be this. Wearing helmets when you’re on a sports bike is very common in the Netherlands. It’s just not common on Dutch commuters cycles, but then again, these do not have this failure mode. Sports bikes in general have much less margin for failure, and thus fail more often and catastrophic than a solid steel bike.
You are very unlikely to crash at 15-20kph and even less likely to hit your head at such speed in a very upright position.
In fact the odds aren't bigger than when running. Do you think runners should wear helmets?
I chose to wear or not my helmet depending on the conditions. Running errands nearby on my girlfriend's lady bike? I don't use it. Riding on mostly separated bicycle paths at my girlfriend's slow pace, same. Riding 3h on the roads or trails on my road and mountain bikes, I wear it.
Typical Dutch commuter bikes are made of steel. And yes, steel breaks, but the failure mode of a solid steel fork is much much different from a titanium, aluminum or carbon fork. They bend, crack and at some point break, but they’ll typically not snap like this because steel is much less brittle than any of those materials.
Also, speed matters. Going 15 or 50 when the thing breaks makes a difference.
Passenger vehicles from this century come with airbags that significantly reduce head trauma. Three point seatbelts have been mandatory in all seating positions for longer, IIRC, and significantly reduce the likelyhood of head vs pavement collisions.
But, if absolute safety is paramount, you do want a five point harness, flameresistant coveralls, neck support, and a helmet, sure.
When you're driving 15 km/h on a flat road accelerating and decelerating at 0.1 g (and almost always only in the horizontal direction) - you don't have the structural requirements for the bike that sport cycling has, yes.
I’m 100% Dutch. I ride road bikes and commuter bikes. I’ve never in my life seen a steel commuter bike break like that. They rust, they bent and they scratch, but they won’t break. Or maybe if you buy a very iffy brand?
As a road cyclist I go 30-50 km/h with a helmet. And as a commuter I’ll keep a moderate 20 km/h without a helmet. Big difference in terms of kinetic energy.
being more likely to catch your fall needs to be computed against how likely the falling is, and how likely injury is. This just confuses things, much easier to look at how many injuries occur.
it's not really a theoretical problem, people actually die or experience brain damage from falling over while standing still; we know a fall of that distance is dangerous. Helmets reduce the impact and save lives and brains. The only real question is the frequency of the falls taking place from bicycles.
Oh, bike helmets. I missed that thread, and I don't want to start yet another endless internet discussion about helmets, but let me point out one thing: the helmet I did wear likely wouldn't have saved my life, had the fork breakage happened one minute later.
> After all, are you strapping on a helmet while you're out walking, or driving a car?
you are much more likely to fall from a bicycle than while walking or getting in or out of a car. For people who are at risk from falling while walking, yes, precautions are taken that they not fall or hit their heads (or hips), it's quite dangerous.
I'm not advocating helmet use. But nor do I advocate for me to pay for patching you up if you don't.
Your argument is that the additional protection of a helmet is not worth the additional inconvenience for you when you are walking or driving.
That's the exact same argument the anti-helmet crowd is making of cycling.
In order for your argument to work, you have to either provide a quantitative basis for why your tradeoff totally is worth it and the other is not, or you have to come up with a qualitatively different argument.
I'm not making an argument, neither about helmets for pedestrians nor cyclists. I'm pointing out that others are not focusing on the right metrics to make their arguments, or are providing spurious commentary.
> the additional protection of a helmet is not worth the additional inconvenience
I actually pointed out that protection of a helmet for pedestrians is called for as standard practice (by others, not by me)
They may help but they aren't really designed for that.
Helmets are designed to limit head injuries when you bang your head on the ground mostly. If you get mauled by a truck or a suv it is unlikely to help much although it will still help if you bang and rebound on it.
When you fall while running, you don't have a mechanical contraption limiting your movement. The nature of riding a bicycle (or motorcycle) makes it much more likely you'll hit your head. People also don't fall as much while running as they do when riding: it's easy to hit a crack in the pavement and lose control on a bike.
> People also don't fall as much while running as they do when riding
Says who?
Just a couple of days ago I had a runner get past me while I was walking on the street, 2 seconds later she hit a bump on the pavement with her foot and fell down hard (she even made a loud splash sound as she hit the ground). No, she didn't have a helmet. Should she have?
At the end of the day, people should be walking around in full steel plate armour, for protection and safety. Or even better, not walking around at all. Bicycles should be banned!
> it's easy to hit a crack in the pavement and lose control on a bike.
No, its not. The person who says this probably hadn't ridden a bicycle in their life, or if they have, not often and not recently.
Bicycle wheels are EXCELLENT at getting over cracks, bumps, ditches, holes, whatever - as long as the attack angle is as close to 90 degrees as possible. It is the riders responsibility to make sure that is the case. Wider tires also help, lower pressures too... and high speeds, counterintuitively.
Of course, if you are afraid of falling or uncertain of your skills on the bike, wear a helmet by all means. Just don't think that it is "normal" and all riders are like that.
>2 seconds later she hit a bump on the pavement with her foot and fell down hard (she even made a loud splash sound as she hit the ground). No, she didn't have a helmet. Should she have?
Did she hit her head? Humans are hardwired by millions of years of evolution to protect themselves in falls using their arms and hands; they generally handle falls very well. Humans have not evolved to handle riding bikes, and using your hands to protect your head in falls isn't instinctive there like it is while running or walking.
>No, its not. The person who says this probably hadn't ridden a bicycle in their life, or if they have, not often and not recently.
This is just stupid and argumentative. Road surfaces are frequently bad and people make mistakes.
>Of course, if you are afraid of falling or uncertain of your skills on the bike, wear a helmet by all means.
Yes, because I'm sure you're such a perfect rider who never makes a mistake.
I came here to say the same thing - "steel is real" is totally true. I did have a Bianchi Pista Concept that was probably my favorite bike of all time, but a no-name steel fixed gear bike that I owned got way more miles and was far more comfortable. The carbon fork on the pista concept helped, but all-in-all it was a performance machine and not a comfortable one.
Anecdotally, and I know the cab material is not the only difference, my RAM 1500 is leaps and bounds more comfortable and pleasant to be in than my F-150. The RAM has a steel body, whereas the F-150 is very light/rigid aluminum. The steel cab does much better with absorbing the energy/sound from the road.
Oh, I wish. Unfortunately a nasty accident bent my steel tube beyond repair due to structural integrity concerns, and I still have yet to find a bike I've felt so at one with... This one was an early 80's Schwinn before the Chicago plant shut down and it's hard to find the exact kind I like.
There's nothing like those lugged Schwinn frames for building beaters. There was a period when some were built in Mississippi, and others in Japan. I've got a couple of those frames, and they're a great raw material.
You can always get a lugged custom steel frame, especially in the US. It isn't going to be a bargain, but it will be specially taylored for your body. Otherwise there's Crust, All-City, Rivendell or other quality brands. Or Surly if you want the cheaper overbuilt stuff. See the links below for inspiration.
I think lack of demand, and a fragmenting of the component standards, are hurdles. Those old Schwinn (and other brand) frames were made during a time period when there was a lot of interchangeability, and most spare parts are still available. Today, standards for things like bottom brackets change every three years. There's no part on my 1985 frame, that I can't replace today, cheaply.
I can get a brand new lugged steel frame that would be perfect for my use -- a Surly Cross-Check. Seriously the only deterrent is a minor injury has forced me to re-think what my future needs are for geometry. I may need to adopt a more "comfort" posture. Better than not riding. Sucks to get old.
Despite my above comment, and thinking about it more, I want my next bike to be modern in the sense of accommodating fenders and much wider tires. I need horizontal dropouts and to be able to set the rear dropout spacing, to accommodate an internal gear hub (IGH).
>Today, standards for things like bottom brackets change every three years.
Your comment is almost a decade out-of-date. Everyone these days has mostly converged on the same threaded BB standard for all but the highest-end bikes. Those press-fit BBs caused too many problems.
> Seriously the only deterrent is a minor injury has forced me to re-think what my future needs are for geometry
A Cross Check with a flat or alt bar and 30 mm of spacers underneath? That or the Soma Doublecross without all the spacers. The main triangle and the fork are build from heat treated Tange Prestige tubes like some of the ATB frames of the '80s and '90s, unlike the Surlys which are built from utilitarian 4130 chromoly. Or the new Disc Trucker with extended head tube and the new Truck Stop riser drop bars. They shortened the chainstays and it no longer rides like barge unloaded, like the previous model.
I went through the same hurdles after an accident some ten years ago, built several bikes in the meantime including a classic road bike with somewhat similar geometry to the Cross Check, which I knew was not for me. Turns out that it's manageable, although I would probably not use it for a 100k trip. My current go everywhere bike is similar in geometry to the Double Cross, but it's aluminium.
I would encourage you to get in touch with a reputable custom frame builder and describe your situation and requirements. The local shop I use has been in business for nearly 50 years, and they use components that I can best describe as "tried and true." A bike they built for me in 2014 is something that I can still fully service myself with commonly-available parts and tools.
Meanwhile I damaged a thru-axle of a high-end big-name brand bike (rhymes with "wreck") I bought 2 years ago, and they told me that they no longer stock that part because it's "obsolete" (!)
Depending on how much sentimental value the bicycle has for you, how much you're willing to pay, and how long you're willing to wait, there are places that will repair your frame. One reason why steel touring/travel bikes are so popular is that you can often find workshops that can repair them when you're in some random city and in a bind. I got a nasty dent in my top tube on one of my ultralight chromoly steel bikes. While the dent didn't compromise the frame integrity because of its location, it marred an otherwise gorgeous bike. The shop that originally built the bike for me cut the tube out, welded in a new one, and repainted.
Oh man I had a bright metallic green Varsity as a kid. That was a great bike and I really regret getting rid of it, or rather abandoning it at my parent's house when I went off to college, and they got rid of it :(
Le tour. The traveler is a great bike too. That settles it, this year I'm going to set aside time to search deeply for a new le tour that speaks to me...
totally totally agree. they were great. my traveler was the same age as your le tour, worked great until stolen from a bike rack by the campus police in a noob mistake, and sold at auction, in 1996 or so, in a big ten campus. i later saw it, totally beat up, locked at the engineering library. (like i alluded to above)
i saved up and bought a titanium litespeed and it's fantastic (for me anyway). i think Lynskey is also great (they're the family that started litespeed, which they spun off to an american investor group, so basically bifurcated with two tennessee-made names for the same group of engineers.
> Your steel fork is bent? Hit it with a hammer until it's true again!
I'm sorry, but this is BS. I don't think you can fix double/triple butted tubes easily, and even if you could, I wouldn't trust them any longer.
I once owned a steel-framed Italian road bike, made around 1985-1990. I bought it cheap because neither I nor the seller knew at the time that it was a real pedigree race bike, which meant lightweight tubes and probably quality steel (I don't know what it was made of, but the whole bike weighed in under 10kg with pedals, which for an older steel framed bike is excellent).
Then I crashed it, I went straight into a lamp post at around 20kmh. It was not a nice experience on myself, but the bike... the fork was bent backwards so that the front wheel went past the down tube. The top tube was bent, dented and bashed and cracked, almost unrecognizably. It was clear from the first moment that no-one will be riding that frame again...
...until a guy turned up to "buy" the frame & parts (I gave them away for free). Guess what, he just needed the lugs and he built a bamboo bike from them! That's how you recycle for good.
Personnally I always felt that the steel is real mantra was just pure ignorance.
There are great bikes of very different materials.
I owned great bikes in steel, alu and carbon. My most comfortable and efficient road bike was made of carbon, the less comfortable made of crude 4130 steel tubing while some other steel bikes were really nice to ride but not as comfy as my current carbon bike. Reason is mostly because these bikes had that old school italian geometry with very short chainstays and couldn't accept tires bigger than 23mm.
I've seen steel bikes bent into unrepairability while I've seen carbon bikes shattered into several pieces then repaired and rebuild with very little difference in ride feeling after the repair for a tiny fraction of the cost a custom builder would ask to replace steel tubes on an existing high end steel frame.
I managed to crack a steel frame mountain bike in the early 1990s - I did use it to commute (about 25km a day) and used it off road a lot. Simply got someone to weld up the crack, got the frame sprayed and it lasted another 5 years or until I could afford to buy something nicer.
Also applies to fixing steel wheels. Can slam them against the ground, hammer them, and step on them to bend them back into shape without losing as much strength as one would on aluminum.
Aside from being absurdly heavy, steel wheels are dangerous since they become extremely slippery to brake shoes in rain or wet conditions. I don't think steel rims are common any more - or at least they shouldn't be unless you use disk brakes or something.
I'm guessing this person might be referring to steel car/truck wheels? A lot of serious crawlers/off-roaders prefer steel wheels because 1. they are cheap and easy to fix/replace, 2. they are easy to weld on beadlocks, and 3. you can definitely repair them out on the trail with a hammer.
Yes, sir, and you had to wait 10 years for it! I've only had steel frames myself but when I try out non-steel frames I can't say I dislike them, just that they feel strange.
I'm still trying to understand the talk about over/understeer and how the bottom bracket is involved. Are you saying you put down more weight on the outside pedal on each rotation of the crank as you're going through a turn?
Also, isn't there some wasted energy going into flexing the frame that should go into the road?
There is some wasted energy from flexing during really hard pedaling, but off-road this can actually make you more efficient when the frame keeps the rear wheel attached to uneven ground.
Have you ridden aluminium bikes in a way that lets you compare? I have a steel and an aluminum frame, but they have different applications so I can't compare. I do feel more harshness in the bars on the ALU frames, like sharper peaks in vibrations.
I'll say it also depends on the fabrication method.
I have an Allez Sprint from a few years ago - it's aluminum alloy with carbon forks+seat, but the joints are welded a few inches away from the joints instead of right at the joint. The result is a stronger bike and an extremely responsive ride that I enjoy for rides <50m miles. You'll feel everything in the pavement, which is both a good and bad thing depending on your preferences.
For longer rides it's hard to beat my old-school fully lugged steel frame on flat routes, but the added weight usually isn't worth the tradeoff for me.
A good alu frame is more comfortable than a cheap steel or carbon frame IME. I think a lot of the perceived differences comes down to tyres rather than frame material - steel bikes are much more likely to have >35mm tyres than alu with 25 - 28mm.
And a really good steel frame (or titanium that has been butted to ride like steel) is even more enjoyable still, at least for a lot of folks. A hand built steel frame is luxurious and will ride like a dream.
• steel, double-butted, Columbus SL from the late 70's/ early 80's
• steel, double-butted, Tange 1 from the late 80's
• steel, not sure but it was a Kona Rova from 2020 I believe
• aluminum, Cannondale R300 from the mid 90's
• aluminum, Specialized SmartWeld DSW from 2016
• aluminum, Cannondale Synapse from the late 2010's
I personally love Columbus SL. I believe it was 0.1 mm thicker all around than even the Tange 1 bike in numbers (but I'm not sure if both triangles were double-butted or only the main one). But it feels springier, even with the same wheels & tires & components swapped between the two frames.
I didn't mind the Cannondale Synapse. It subjectively felt better than the Smartweld aluminum, but the difference could have been in the wider tires on the Synapse (versus the Spesh).
The Kona Rove... for some reason was a bit of a letdown. People seem to love that bike, though.
And Hi-Ten is as stiff and clunky as everyone says. Pure beater material right there.
But the Cannondale R300's mid 90's aluminum was even harsher and wrist-numbing. It'll rattle you to your core!
All of this is anecdotal & subjective, though. I've learned to compensate the added road-noise from the Specialized aluminum by hovering off my saddle and bending my elbows a bit while loosening my grip.
But man, I love that racing steel bike. It makes me wanna try Tange Prestige of the road variety, and Ti as well.
Only other thing I'm super curious about is CAAD 12 Aluminum, but on forums I've read that CAAD 12's are slightly harsher than SuperSix Evo's (all carbon).
- Steel Pogliaghi Track bike. 80's vintage, handbuilt. Thinwall standard diameter tubing. Probably the flexiest bike I've ridden, super easy to see BB and fork flex.
- Handbuilt steel road, mid-90s, Columbus OS tubing I think. Kenesis bonded AL fork. Pretty harsh. The carbon forks of the time were much smoother. When I get it from storage (where it's been for a decade), I'm going to put a custom steel fork on it and ride it on sunny days and smooth pavement.
- Handbuilt steel 26" road tandem. Late 90's. hand built steel unicrown fork. I _love_ this bike. It's running wide Rene Herse tires now. It is so planted on descents with bad pavement. My next single is going to try to replicate the feel of this bike in a single. (probably something like a Crust Lightning Bolt, if I can get one in the EU)
- An Al Redline Cross bike (kinesis unicrown alu fork, which was an out and out noodle), and later with a Surly crosscheck fork (harsh, heavy). Very comfy till I killed the fork. Sadly, there's a fatigue crack at the bottom bracket.
- An over stiff Al gravel bike (PlanetX Full Monty) w/ carbon fork, mildly redeemed by 650cx48 RH tires. It's so stiff that standing really isn't encouraged, the bike just feels dead.
- Inexpensive Carbon road bike (PlanetX pro carbon) with carbon fork. Nice enough for the 6k miles I used it, but I'm done with 25c tires and no space for fenders. It's no faster than the gravel bike, despite being 4 kilos lighter.
- Mid 90's Cannondale 3.0 frame. Stiff. I greatly preferred the Steel one that replaced it.
My experience would say -- The forks really matter. The lighter unicrown forks are so much better than the super stiff ones. Wide, supple tires matter too. They're a good 7+% faster on the tandem, and probably the only reason I like riding the AL Gravel bike.
I have two keep-for-life bicycles made from Reynolds 853 steel and both are exquisite. I'm not sure what the modern equivalent is. But I need to find out soon-ish. I need one more nice touring/gravel bike for my final two decades of riding.
That's great to hear. I kinda wanna get my hands on a LeMond Zurich (or better) in my size in that fabled 853 if it ever pops up. Got to test ride one super briefly and I immediately liked it.
I've seen some seasoned looking folks on Saeco's blowing by me and my buddies on our more modern goods. If that kinda nostalgia coupled with the lore of steel I'd be all over it.
From the Prodigiosa (part of Gios which still makes steel bikes following their philosophy from the 70s-80s bicycles) website[0]:
> - The true quality of a frame is felt in the descents
- In a racing bicycle, half a degree in the construction of the frame is more important than half a kilo in its weight
- 20 grams on the wheels are more important than 500 grams on the frame
- The tubes of a frame are like the ingredients of a dish. All are important but the taste depends on the skill of the chef
- A racing bicycle must be made to measure
- The bicycle is the heritage of Italian Culture, and we must preserve it
- A steel frame is like a gold coin. It keeps it's value as the years pass by
- The frame is the heart of the bicycle, and the groupset merely the clothes it puts on
When I was working in a shop I had the opportunity to ride a lot of different bikes. Generally steel felt less harsh which allows for longer rides and tracked significantly better. Although this was a couple decades ago.
There's more nuance to that. A 4130 steel frame with oversize top and seat tubes and thicker walls (0.9/0.7/0.9 mm) typically used on touring bikes is also going to be stiff. A too thin down tube will make the bottom bracket flex too much.
It's almost impossible (for any of us) to do an apples-to-apples comparison, because there are a huge number of non-frame-material variables: tires, wheels, bars, seatpost, fork, geometry, etc. etc. Some of these factors (e.g. tires) make a far greater difference than frame material. Even if you could isolate the frame material, different materials can be tuned in different ways for different purposes. I therefore don't think it's worth focusing much on the "feel" of different frame materials.
> Have you ridden aluminium bikes in a way that lets you compare?
I rode a boring old aluminum 6061 GT for a couple of years back in ~2010, and I now own a Cannondale CAAD12 aluminum alloy frame. I ended up putting it full-time on trainer duty. I thought it was stiff, and the overall "feeling" was crisp. If I were feeling uncharitable I might say it's harder to control. I hear the CAAD13 frames improve on ride quality.
I've found it very difficult to find a non-steel bike that offers the same ride characteristics that I've become accustomed to. And honestly at this point I don't have a lot of time and incentive to keep trying different things when I feel I've found something that just works really well for me.
In a way, it was, because they could go super large tubes without the weight penalty. Kleins and the Cannondale 3.0 and 2.8 were the first to really crank the dial on stiffness, when a lot of the other bikes were running normal or +1 oversized tubes.
They were super harsh because they were aiming for stiffness, and the material allowed them to do that.
Yes but are dispersing power. It’s the same (endless) story as the vinyl VS digital… you can “feel” better the vinyl but the digital is technically better.
I tend to agree - unless you had the ability to compare the ride of identical frames made of different materials - same angles, same components and tires, I suspect a lot of the 'feel' of the ride comes from having been told steel is more forgiving, aluminum is harsh, etc. We have a lot of very subtle control over the frame, angles, wall thicknesses, and I would expect a skilled builder to be able to produce the desired performance for whatever frame material. That said, there do seem to be some obvious poor choices, I am not going touring on an all-carbon frame that could fail catastrophically in the middle of nowhere, no matter how comfy.
Depends on your goals though I think. Since ride quality and feel is a factor, it's not objectively true that carbon is better than steel. Just that it has better stiffness. Do you want better stiffness? Well it depends entirely on what you are doing and what you want to experience.
For example, in a touring bike, you might prefer steels robustness and softer feel for long hauls. Some people might prefer carbon for it's lightness. It's all down to choice, neither is objectively better.
Yeah absolutely correct but the article was talking about professional cycling. I’m also a cyclist (not touring oriented) and the stuff a bike is, the more I like it.
A racing bike should flex vertically but not horizontally.
Where is the power dispersed? Typically bike frames are made of elastic materials which return the power that was injected through the deformation. If the power isn't returned, then the deformation was plastic and you have a bent frame, so that isn't really an option.
Yes but it returns the power when your pedal is at the “death point” and you aren’t making power on the frame, then the power is dispersed through the frame flex when not needed. The frame should be soft vertically on the seat tube.
I'm not sure if I see it. If you're not putting a force on the frame, then the frame is not twisted, and it doesn't return anything. You have to push in order to recuperate the energy.
If you remove the force so fast that the frame doesn't have the time to return the energy back to you, then it will disperse the energy by ringing. But i don't think it's realistic unless you drive your pedals with hammer strikes.
I would think that most of the energy goes in between 12 and 3 o'clock and then flexes out from 3 to 6. So if the flex pushes back against your leg a good part of the energy should go back on the road.
The points the article makes about torsional and lateral tension helped me understand better why I prefer steel over carbon. I've always described it as "feeling more in-sync" with the bike. For example, the subtle instinctive biomechanical action of putting more weight down on the outside pedal to re-align the wheels when you start to understeer just feels great, like you're "one" with the bicycle.