Their modeling suggests
Psh, it's only a model.
We know proto families and such exist going back through to PIE and others. We know that languages split off of those and, if I'm remembering correctly, proto-Germanic itself would hit in their 0 +/- 1000 range easily.
After we domesticated plants and animals, however, the planet’s population ballooned—and so did the number of languages.
This feels counterintuitive to me. We had small groups of hunter-gatherers that interacted with other groups far less often (at least presumably) before agriculture. If people generally don't speak like their grandparents today, why wouldn't that be the case then? I guess one could argue that fewer people meant less drift, but i don't know about that.
After agriculture (and arguably even before it in some places), we had more permanent societies with structure, population, governance, etc. Why would this be the place that language drift/division starts happening rather than the opposite?
I'm not a trained anthropologist or anything; it just feels weird to me.