Can anyone give a road map for how Bhatt–Scholze's fancy recent p-adic work applies to questions in more general algebraic geometry and commutative algebra? I'm aware that it does, following Andre - La conjecture du facteur direct, but I still don't really get how the full picture works.

My question is simple, but broad:

How and when do broad problems in algebraic geometry reduce to p-adic problems which can be attacked using these methods?

Moreover: which problems can't be attacked this way? Why?