Have you read Bryan Caplan's Selfish Reasons To Have More Kids? He makes a pretty strong argument that middle and upper middle class parents wildly over estimate the effects of more time and money spent on their children.

I have heard Caplan talk about it with Russ Roberts. I have no doubt that Bryan Caplan is right at some level of detail, that there are plenty of things that any given upper class parent does that are less effective than others, that as with so many other production processes, there are efficiency gains to be had by bringing scientific approaches to studying the production.

What may be missed is the net. What is the net effect of rich people grooming their kids compared to the baseline of poor children being ignored and only beaten a little, and living in this "good enough" environment you get for almost free? You say nepotism, but I am sure that for 99.999% of jobs I have no idea whether the applicant behaves well because they are actually linearly descended from one of the 100s of millions of people I'd be presumably class allied with, but I have every idea whether their grooming, speech patterns, ability to self-deprecate, ability to approach tense situations with humor, calm, and even deference rise to the levels I am concerned about. I have every idea whether they become defensive when I am testing them. In short, what looks like nepotism to you is primarily me looking for the mix of characteristics that I understand are needed for humans to cooperate deeply and pervasively in a way that goes orders of magnitude what are genes without culture ensure.