I think a somewhat neutral summary (of someone still annoyed by Vercel/Next) would be like this (Notice the distinctions between Site and App, not always clear cut but a dividing line imho):
- React was created by FB to solve real technical issues as their frontend became larger and more complex.
- Site creators liked it as it was one of the solutions of a real issue of reconcilliation of state and view (that often wasn't so bad in the big picture) but React was often a bit heavyweight, App creators really loved it as state reconcilliation took away that entire class of bugs that just became so much worse quickly as Apps grew (and React allowed for more people to create larger apps).
(Angular and Vue has always done this also, they are parallel developments)
- Pressure from those doing sites has always pushed development of React to be "simpler", often good for most parties (even if I think that Redux was mostly thrown overboard prematurely).
- Part of simplifications was bootstrapping, create-react-app became one of the recommended ways to start projects (and was also incorporated into other toolchains such as .NET templates)
- Heavy builds, disabled JavaScript and SEO issues was teetering issues (especially for public site builders), not entirely sure of the inspirations but Next did solve that (perhaps not always entirely elegantly initially)
- React internals start to change to better support these scenarios, nobody really has objections since changes in React has seldomly been for the worse (functional components, hooks, etc). Vercel gains traction as a "do-good" choice.
- After all troubles of OpenSSL, Node finally adopts OpenSSL 3.0 thus breaking create-react-app that had been "deprecated" by the React team (it's easily shimmable but it sent people looking).
- People looking for options find that the only "official" way to use React according to the site is to use Next, so many start adopting it out of fear of being left behind again.
- The Next model however is quite different and tailored to "site" builders and/or people running the full stack in JS
- React however is quite popular outside of the JS only world for enterprise SPA and/or mobile apps where trying to shoehorn in a Next "frontend-backend" becomes overkill and extra complexity. (We used it for one or two projects but have now abandoned it for our regular work).
- The React site is updated slightly, Vite and similar are now mentioned but the perception damage is there and hasn't let go (and last I checked using f.ex. Vite was not "recommended" as being an inferior option to Next for React usage)
- A very popular option for CSS-in-JS (styled) becomes deprecated due to React internals changing for Next and requiring significant rework that the original author had no interest in (no really clear successor with support across the board for Next, SPA and React-Native scenarios hadn't appeared last we checked).
Now this is my perception of events and I'm pretty sure that I'm not alone in this, the Next/React authors felt like it was the way forward due previous feedback for those that hurt (site builders) but probably misjudged or didn't appreciate how much React was used in other workloads(apps) that got disturbed while they were improving their thing.
That Vercel has managed to alienate people in other ways like billing (or politics?) certainly doesn't seem to have helped either.