Doesn't seem to be mentioned in any other comments or the paper itself, but this is Berkson's paradox.
Could be. But then, it's still a reasonable heuristic, because you might find that unattractive managers are uniformly pretty good or better, and attractive ones are 50-50. The best manager might be one of the really really good looking ones, but overall there is less selection pressure on attractive managers with respect to performance. (Halo effect?)