It's not 129,000% markup. You're paying for outgoing bandwidth (at S3 prices) and you're paying for the Kindle's incoming bandwidth, which could be over an internationally-roaming 3G network. Yes, it's dumb that Amazon charges authors and not themselves, but would you prefer they take 40% instead of a fixed $2?
Realistically, how many downloads are really done via international 3G roaming? Much less than 1% I would suspect. And while it's true that there are some plans with absurdly inflated roaming costs that could still drive up the overall average to $2+, it would be beyond incompetent for Amazon to agree to such plans for their Kindles.
One factor that nobody seems to have mentioned so far: was this massive delivery fee in any way knowable to the author beforehand? I'd consider it a bit disingenuous of Amazon if they hide it.
My guess is that you're not paying a lot for each bit transferred internationally, you're paying for the ability to be able to download content internationally. (Authors may be subsidizing the devices, as well. It makes sense, though; if Kindles were less ubiquitous, how much money would you make selling e-books.)
It's kind of like complaining that your ISP charges you the same monthly fee even if you don't download anything. The charge is for having the ability to transfer bits, not for each bit transferred.
I guess the complaint is that Amazon transfers much of their financial risk to authors via a $2 delivery fee. All I know is that I wrote a traditionally-published book and saw about $5 for every $40 the publisher charged. I would gladly take $38 instead of $5 and they could call the fee whatever the fuck they wanted.
