You're actually saying the same thing I did, Dave, with only one difference I can perceive. You're making the assumption that the parents are not part of the problem. In many cases - observable in public places - they are. So, if a series of teachers (doesn't take many - just a few in a row), some activity leaders, and the parents all fail to let them experience failure, reward them for any level of participation, etc., then that may be enough to dilute the effect of more useful reward systems. It doesn't have to be universal. And it is harder to fix once the child has begun to build a paradigm based on those rewards, so what happens at 10 may not be as important as what happens at 6.
EDIT: Oh, and the "hard austere" bit serves as a strawman here. Nobody suggested that model.