This article makes the mathematical point, but deliberately ignores social reality. Yes, 100 generations is sufficient time for the premise to be possible, to

This article makes the mathematical point, but deliberately ignores social reality. Yes, 100 generations is sufficient time for the premise to be possible, to fully propagate genes from every ancestor to every descendant. But reality doesn't work that way. Human population groups breed within themselves. Mating pairs are not chosen at random from the entire human population. Consider all the remote groups that would have to be covered to satisfy the assertion. Someone alive today would have to become a common ancestor of every Lapp in Finland, every Aborigine in Australia, every Inuit in Nunavut, every Maori of Polynesia, every Falklander, every North Korean, and countless more groups that barely have any contact with the world community, let alone interbreeding. (And maybe we'll actually launch a Mars colony or interstellar generation ship, literally making it impossible.) The article does mention this very offhandedly, inserting the clause "If people in this population meet and breed at random". But that antecedent is plainly false so the conclusions are not defensible.