I think the theory is that for code review to be its best, it sits on a contextual pyramid composed of reviewed design docs, pair programming, standups, and all the rest of it. That the "reviewers" are not hearing about the background, motivations, alternatives that were considered, broad implementation choices, etc etc for the first time when the PR goes up.

But often there are some or all of those pieces missing, and so yeah, code review ends up being a bit shallow and performative.