I think this analysis misses the mark. My opinion is that there is a serious tension between some of StackOverflow's objectives (i.e., to have some "Wikipedia-like" "wisdom of the crowds" effect cause answers to converge towards the ideal) and the incentives they have put in place (which tend to reward the first semi-correct answer over later, better answers). It will be interesting to see if some form of appropriate tweakage can diminish the gap.
