I can't help but to read this as "guess the omitted variable". There's just no way of controlling for everything that might explain this, and it's obviously not the grandparents' time in educational institutions in itself that does the trick.

Thankfully, one of the authors summarized it well:

This opens up a myriad of possible explanations and will need to be replicated.

It's easy to imagine ways this effect makes perfect sense, especially if it's small. So the question kind of becomes what they have managed to successfully control for.