I think PRs are also the wrong place to litigate any of this. Requirements, specs, architecture, etc sure. But most systems I've worked on that were especiall

I think PRs are also the wrong place to litigate any of this. Requirements, specs, architecture, etc sure. But most systems I've worked on that were especially BAD were that way not because of some PR-reviewable syntax or style, but because the whole foundation was wrong. The most idiomatic, consistent, linted, unit-tested, etc code in the world doesn't matter that much if the underlying architecture is a mess.