It can be a good demotivator when the disparity is large and hard work seems unrewarded, though - it can nurture a fatalistic outlook in which the rich are rich because the game is rigged, they "chose the right parents", etc., rather than working hard to get there. The rewards have to seem attainable.
That's exactly what pg was referring to when he spoke of historical contexts and the controls other than the market in allocating wealth.
There have been some very interesting studies done on this by a Stanford professor with children from the projects. (I'll look it up if you're interested.) The feeling of despair and hopelessness sets in very early.
I understand the theory you're presenting; however, I have found the opposite to occur in reality.
There are a lot of assumptions in there that some people don't like to talk about. Fairness of the pseudorandom function and other forms of motivation are obvious ones, but there are many more. It's still a credible argument for why disparity in wealth could possibly be a necessary consequence of something good. It's just not a credible argument that this is necessarily the case.