Thats something a have regarding PR and code reviews: much frustration can be avoided by discussing what to do and how to do it before showing the work done (avoiding also infinite bikeshedding). Often code review talk about the strategy to do X and not the way to do X