All disagreements are about "weights", and it's instructive to ponder what this one is about. IMO giving people a process/documentation does help to some extent

All disagreements are about "weights", and it's instructive to ponder what this one is about. IMO giving people a process/documentation does help to some extent, so the dichotomy the author presents here is between those who think doing so mostly solves the problem, and those who think most of the way is still left to go: so this one is I guess about the weight one attaches to the question of how much the non-technical solution solves the problem / how much you can rely on people to do what you'd like them to do, etc. And while experience may often involve lowering that weight over time, it's also important not to get too cynical (set weight to 0), and continue to attempt to estimate the parameter accurately: maybe the "ask people to do X" solution will give you a 30% probability they'll do X in any particular instance -- then ask how ok that is, etc.