The extremly high diversity of indigenous American languages is quite perplexing for me.

We know that indigenous Americans didn't have high genetic diversity. the first humans entere there before 10000 BC and the majority of them where Q-M3 haplogroup. Also some C3b and perhaps R1b. So at the starting point they had 2-3 languages then later we should expect the same number of language families.

But the end result is very different. More than 100 language families (!!! not groups). Only in California there was more than dozen of language families.

How this can happen?

I think the reason is that when those people entered America they didn't have mature and stable languages. They have perhaps some basic vocabulary ( less than 50 words ) but not fully developed language. The human languages appear after 10000 BC so in Indigenous America we see the first burst of this creation process.

Perhaps this may be aplied for languages in Old World too.