Kenpom's latest blog entry covers the late game 3-point situation and what's a smarter option: fouling or not fouling. For those of you still arguing one way or the other, it's a really interesting read. If you don't want to take the time to read it all, the basic conclusion is that losses are so rare that neither option is more of a guarantee than the other.
He does note that it's possible that Michigan, by virtue of having fouls to give, might be in a slightly different position than most of the games in his analysis (in fact, the Michigan game doesn't fit his sample set, as he chose games where the team that's losing gets their possession with 5-12 seconds remaining, presuming those situations lead to a better quality shot than what Brust had).
Personally, I'm in the "shit happens" camp where you play defense and make them hit an extremely low percentage circus shot.