A software debate heuristic I use is “is it easy to change later”. If it is, then most likely it ain’t worth arguing about too much. Argue about the things that relate to strategy.