Cool.
Do the models fit real-life data, especialy regarding phenotype?
I don't know. I've never done that. There are some studies (which I read, but I don't remember the authors; it was some time ago) that trace gene frequencies using stochastic differential equations. Perhaps a search of GOOGLE SCHOLAR would be useful.
A simple application would be the proof that there is always a "least recent common ancestor" (African Eve). The relation "mother of" (applied to females) is a contraction mapping (there are always fewer mothers than daughters); and thus if iterated, there will be a fixed point; being discrete, the fixed point must consist of a single individual. Similarly for males. Similarly for names.