I am not totally sold on the argument that Natural Selection's merit function is purely to maximize offspring. It is trivial to come up with examples where that is a bad strategy. For example the smaller a herbivores ecosystem, the less they reproduce. A species which reproduced at the rate of locusts but only ate rare amazonian berries would very quickly drive the berry to extinction and then itself. So a more accurate "goal" for natural selection might be to maximize long term, stable, populations of offspring.

This ends up being kind of important, since it starts to explain behavior which otherwise doesn't make sense in the short term. E.g. It has been shown quite convincingly that altruism is a heritable trait in genetic systems. Essentially, even if X% of your tribe has the "altruism gene" it can be beneficial in the long term for that gene if you die heroically protecting your tribe.

So I think there might be a convincing argument there that if natural selection / evolution is pushing us towards stable long term offspring population. If so then our sub goals may not be that far off from optimal.

idk, more analysis needed lol