These are strawmen arguments. No one is arguing that libraries matter more than organizations/teams, and no one is arguing that it's impossible to build good/ba

These are strawmen arguments. No one is arguing that libraries matter more than organizations/teams, and no one is arguing that it's impossible to build good/bad frameworks in any given language/framework/etc. You can (and should!) solve the problems with your teams and your tools. If you pick a tool that solves problems you don't have (probably because it's new and shiny), there are probably hidden costs that you didn't consider and you're probably going to slow your team down. How much that slow down is going to affect your project depends on how significant the slow-down is and how sensitive your project is to developer productivity (the widely used internal tool with the big budget at BIG CORP is probably safe, but the team at the small startup with the shoestring budget and fierce competition is going to feel that friction in a bad way.