That's the problem. Most 'people', such as budding developers, do not want to wire-up the internal components. They don't want to think about the complexity.
Until you need a tool that your current toolchest doesn't offer, you're fine, and likely benefited from having simple, easy-to-understand tools. Switching to a different toolchest will not be trivial, but it alters which problems are easy, and which are hard.
I understand what you're getting at, but there is real architectural choices right now that didn't exist in 2012. It's true that React enjoys good mindshare and attracts new talent based on name-recognition alone, but there already is a next wave of alternatives to React (like Vue, Mithril, Aurelia) that are borne out of lessons learned from React and Angular/Backbone/Ember. If someone discovers that React doesn't in fact meet their needs, there's others to turn to, organically.