While I agree that a major strike would be horrifically catastrophic, I think it's easy to underestimate the platform for life that exists on the Earth. Unless the biosphere is entirely wiped clean, it's always going to be easier for us to persist on Earth than in space, because this is where we evolved.
We don't understand our global ecosystem well at all. Think about one tiny aspect of human life--the gut bacteria. We are starting to learn that it is important, but we don't really know for sure where it comes from, what it does for us, and how changes in its makeup affect us. The Earth is TEEMING with bacteria, most of it unknown and unstudied. We have no idea what the long-term effects would be of removing ourselves from that permanently.
And if you can think of any risk to the Earth, it's greatly magnified for a spaceship. A comet might hit the Earth? Well even a small meteor will wreck a spacecraft. And there are a lot more of the small ones, and they are harder to see and track.
In the very long run, I hope we do figure all this stuff out and become a space-faring civilization. But in the short run there probably needs to be other reasons to try living out in space.
The two that I can think of are tourism and energy research. Tourism is a huge business and if you can imagine a luxury hotel in low earth orbit, well, it will need a staff and logistical support. As launch costs go down, this might start to make business sense.
For energy research, a multi-terawatt solar array in zero gravity might let you do all sorts of interesting things with exotic materials and high energies, like work on manufacturing antimatter.
I totally agree that most catastrophes (including anthropogenic ones like global thermonuclear war, unstoppable climate change, etc.) would be extremely unlikely to destroy all life on Earth. But destroying humanity would take quite a bit less, so I am willing to entertain arguments for colonizing the inner solar system.
I love end of world books, and I had only read one Stephenson book before, the Cryptonomicon. I was not ready. Eventually I thought, "okay, this must be almost over and can't get much worse."
I wouldn't say it's terrible. Instead I'd say it was deeply flawed and I found it profoundly disappointing.
SPOILERS HEREWITH
Also, big brain dump.
The fundamental premise makes no sense --- the amount of energy needed to blow up the moon in the way described is about 10^29 joules, if you do the maths, and even the sun takes six minutes to release that amount of energy. It's not survivable to be in the same solar system as that, and I'd be unhappy about being in the next solar system; I'd expect some high-fraction-of-the-speed-of-light fragments coming along fairly soon. The first question any sensible physicist is going to ask after seeing the moon explode into fragments like that is 'why hasn't half the Earth been vaporised?'
(Not to mention that there's no mention as to what the fragments of the moon are orbiting around.)
The research is poor. When the Hard Rain comes, anything in low orbit is going to be toast, because all the bolides are going to have to pass through there to reach the Earth. They literally build the Cloud Ark in the worst possible place. (Why not boost each piece out to geostationary orbit, or better, the same orbit as the Moon but on the opposite side of the Earth? Any fragments you meet there will have very low relative velocity, so you can assemble the habitat there in safety.) Likewise, the entire concept of the Cloud Ark can't work --- it relies on burning thruster fuel all the time, and there is no way to replace this. It's fundamentally unsustainable.
(Also, it suffers badly from Darlings: this is when an author discovers a Cool Thing, and is determined to use it whereever possible. In Seveneyes' case, it's whips. When an editor tells you, kill your darlings, this is what they mean. Unless it's important to the plot, this sort of thing is simply an indulgence and should be removed.)
The plot is... very telegraphed. Very, very telegraphed. He made such a big thing of the mines and the submarines that it was painfully obvious they were going to important later. Indeed, in the last section I found myself saying, "yes, all right, the people from the mines are coming up, JUST GET ON WITH IT AND MEET THEM". And then when the ex-President shows up at the very last moment before the Earth gets destroyed, the entire remainder of the first-section plot flashed before my eyes, and it wasn't a good sight. (Particularly after the promising beginning when he opted out of political soap-operaing entirely.)
Then there's the weird tonal shift half-way through the book, when the cannibals appear. At that point, there is no possible way any of them can survive; they simply don't have the population any more to maintain the skills to run the equipment. As Stephenson says, they live in a high-tech incubator; they need people, lots of people, because they need knowledge to make everything run. With only a double-digit population they might as just give up and die. Instead we take a hard left into some weird metaphysical area, and then whats-her-name gets her magic gene machine out and... I don't know; I had no idea what he was trying to do here.
Then there's the last section, set in the far future. We get a bit of closure from finally seeing the other end of the submarine and mine plots (yeah, and what happened to all the other submarines? And all the other mines? Oh, and his explanation of why the mines couldn't expand doesn't make sense either), but it opens up lots of other plot threads, and seems strangely disconnected from the rest of the book.
As you can tell, I feel quite strongly about Seveneyes; parts of it are really excellent, especially the first third; so much so that discovering that the whole thing was just sort of... incoherent and rambling... came as a horrible disappointment. I don't know what the book is trying to say, or even if it's trying to say anything at all.
Oh, well. I'll just have to go reread Anathem now.
Also, it suffers badly from Darlings: this is when an author discovers a Cool Thing, and is determined to use it whereever possible.
Sir Arthur Quiller-Couch meant something else when he said "Murder your darlings" in a lecture -- later published in 1914 as "On Style" [0], the final chapter of his book, "On the Art of Writing".
He states that style is not about "extraneous Ornament" and gives an example. And at the end of that paragraph he goes on to say:
...and if you here require a practical rule of me, I will present you with this: ‘Whenever you feel an impulse to perpetrate a piece of exceptionally fine writing, obey it—whole-heartedly—and delete it before sending your manuscript to press. Murder your darlings.’
In the c2 wiki entry for "Kill your darlings" [1], the point is made that this can apply to programmers too:
In software design, when you find yourself feeling particularly proud of a neat little bit of design or code, stop and ask yourself how someone who didn't give birth to it will regard it. If it turns out to be overwrought or too slick for the need, you should probably kill your darling and replace it with an ordinary solution that others can actually use, and not just marvel at.
(Also, the word "glitter" is used a bit later there to describe the stuff that a writer should get rid of, which is wonderful, since it evokes the (relevant!) saying, "all that glitters is not gold" -- and because the point is made that glitter can be distracting to the reader, who is not as maternally invested in it as the writer.)
The c2 wiki page also mentions Samuel Johnson's related phrase: "Read over your compositions, and wherever you meet with a passage which you think is particularly fine, strike it out."
Until we are a space faring civilization, we are one asteroid/comet strike away from extinction.