Neat! I did not think of generalizing my arguments; we could call it the resource commit fallacy. We need techniques to help us solve this problem.

One strategy that comes to mind is precommitting to allocate the resources to the most efficient place before you optimize yourself. Then taking a bet with someone that if you fail to follow through with your commitment, you take a penalty of resources.