Bill it’s just a rate of fixation of neutral mutations, which would be equal to the rate of occurrence in a population of constant size.
So for chromosomal fusions (which we know empirically are mostly neutral), you just take the rate at which they occur and then the mean rate of fixation is the mean rate of occurrence(there would be some variation around the mean of course). Same would be true for gene gains and losses(duplications and deletions/pseudogenizations) which generally have very small fitness values in the small effective populations of large animals.
So you look at the phylogeny, see that a handful are different between individual deer species. Can a handful of those occur and fix in the couple of million years that separate them? Yep.
That’s it. That’s the “population genetics model” you’re asking for. The End.