> I also suspect that personal preferences like that aren't all that sticky across more than a few generations

There is pretty strong evidence that the variance in the number of offspring is about 50% genetic.