I think an even stronger statement holds true: if you don't actively redistribute then you're effectively making an open-loop amplifier which will amplify noise (who happened to be in the right place at the right time) in preference to signal (effort input) -- which is just as harmful to the cultivation of a beneficial incentive landscape (i.e. one that promotes effort input, since that's tautologically all anyone ever gets to decide) as a failure to reward success would be.

Alternatively: it's often better to reduce the transition-state energy than to improve the final reward, which is exactly the opposite of what happens when decisions are made by people who have 1 metric shitton of money to invest in building a moat with the aim of returning 2 metric shittons of money.