Economics in One Lesson: http://www.hacer.org/pdf/Hazlitt00.pdf

It's a great read. It's brief. It doesn't go into a lot of technical detail but it does illustrate a central fallacy that many people, even some economists, make when thinking about economics. If you haven't read this book, I'm going to assume you're ignorant about economics until proven otherwise.

I don't think attracting late adopters will be an issue. If Bitcoin survives it's early growth/volatility phase, the late adopters (who by definition are always late to the party) will get over the "it's new/scary/phony/scam" and start thinking "it's normal".

3) The precedent they're setting here cannot be maintained.

Exactly what happened here may never be fully known, but according to Mt Gox an unauthorized user accessed someone's account and placed a sell order. Passwords get guessed/leaked all the time, and any exchange that attempts to undo that in every case will undoubtably fail.

If I'm careless with my password and someone places orders on my behalf in my account without my permission, will Mt Gox revert an hour's worth of trading to fix it? What if I only had 2 bitcoins? Or 20? or 200? There is no way rolling back trades to handle a compromised password in any way that will scale to the size of bitcoin's current economy. Unless Mt Gox is wiling to explicitly say they'll give this same treatment to any user who has their account compromised, it's blatantly unfair to everyone else.

This also opens the door to allowing anyone to request equal treatment if they made some trades they later regret. Log in through a proxy to make it seem like someone from a distant country was using your account, make your trades, then later scream about how your account was compromised and you want a do-over.

Wow... because I'm a noob.