That's a good point about micro decisions. I remember going to an excellent "software carpentry" course, where one (of many) really useful tips was to think of it not as "top level spec, then down to the detail and stay there", but "keep popping your head back up to look around at the big picture."
It seems to me that as well as diversity as a tool for helping to get balance in those micro decisions, various techniques can be used - like coding in pairs, frequent code review as an ongoing process (I just did a preliminary code review of a colleague's stuff for this project, and it was good to do it part way through rather than at the end, because he's now gone away to take on board a lot of the stuff I raised and write the next chunk of code in the light of the review - I think it's really changed how he's thought about it, introducing the idea of thinking about it from the point of view of the end-user coming to the completed code and accessing it at any random point in the suite of programs rather than the coder writing it sequentially. Conversely - because I'm a scientist, rather than a coder - I've learned a lot about better techniques for the nuts-and-bolts of coding from doing the review, so hopefully I'll write better code).
I agree it's disappointing that we have to appeal to people's mercenary, goal-oriented side to get equality taken seriously, rather than them seeing the inherent unfairness and wanting to tackle it just out of a sense of what's right - but if the end result is a win, I'll take it!