That brings some tweaks and ideas to mind, but I obviously need to take a long break and do some serious reading before retrying my hand at amateur population genetics. Any resemblance to useful data in that mess is entirely coincidental.

It would be interesting to make some plausible adjustments to see what happens to strictly male-line inheritance. One could substitute probabilities for the strictly 10/80/10 divide I used and make plausible assumptions e.g. probability(male is dominant|male is descended from dominant male)>probability(male is dominant|male is descendant from monogamous male). But I'm betting this sort of thing has been done elsewhere and that the job was better than that.