dcs, you’ve had incomplete lineage sorting explained to you at least three times (that I am aware of). Those explanations seem to have had no effect on you.
Have you ever wondered why the pattern of relatedness between humans and other primates exactly matches the ILS pattern? Humans are the closest relatives to chimpanzees, then gorillas, then orang-utans. This pattern of relatedness predicts that humans will share the most ILS with gorillas and then orang-utans. Indeed, we were able to predict in advance how much ILS we expected in both cases, and the predictions matched up with the observed values very well.
As for probabilities, it’s not just mutations in two species you have to account for (not that your “explanation” for two species works). We see scores of shared mutations across many species (for example humans, chimpanzees, gorillas and orang-utans. Do you have a sense of how improbable it would be to have all of these mutations crop up independently in unrelated species, yet give a nested hierarchy pattern? Every mutation event is an extremely low probability event to begin with.