I'm starting to think that code reviews are more about showcasing one's own knowledge of obscure edge cases rather than actually improving the code, and it's a huge waste of everyone's time - can we please just focus on making the code work and ship it already?