can we please just abolish "nitpicking" from code review? you know, the kind where someone points out a typo on line 37 of a 1000-line PR but doesn't actually comment on the code's functionality or correctness?