IMO it's a useful word that accurately describes what happens to a lot of tech companies that grow too quickly (or try to, at least): They abandon their early/original customer base in pursuit of higher-profit customers. In the process, they often modify the original feature set or UX or pricing scheme, etc., in a way that negatively affects their early adopter user base. It's only natural for those early users to feel abandoned or betrayed.
Just blaming it on "market forces" seems like a bigger cop-out than actually calling out "enshittification". Companies have different cultures and values and leaders and different kinds of investors, and they can lead to different growth models and customer bases. It's in particular the kind of company that seeks fast, explosive growth (VCs and founders looking for a quick exit, maybe?) that's most likely to enshittify. Small, slowly-growing/mostly stable companies usually don't enshittify as much... bu you've probably also never heard of most of them.
-----------
Vercel was once an indie darling too. I still wouldn't call it "enshittified", but their focus has definitely shifted. The rendering model of recent Next.js is way more complex (powerful too, but a lot harder to reason about) and Vercel the company is trying to value-add a lot of things (preview comments, toolbar, etc.) and also diversify into more backend/CDN/DB/KV stuff, like Cloudflare's other offerings. I still prefer Next & Vercel today, but I can't say people like me (small time web devs whose customers are mostly small businesses) are their primary focus anymore. What used to be a really amazing frontend web host for simple SSG sites with easy caching is quickly trying to become the next full-stack web monolith, and the number of times I've run into Next issues on other hosts over the years has led me to not even bother trying to deploy it anywhere else but Vercel these days. That was probably their end goal all along.
Whether it's an "enshittification" is a judgment call, but it pretty unmistakably feels like a deliberate pivot towards the enterprise and "real" web use cases that have a lot of complexity and need all the power of mixed static/server/client component trees. In response, I see a lot of simpler frontend projects move towards Vite or Astro instead.
I'm still grateful for Next/Vercel to have unified so much of the JS world that was previously super chaotic before Next took over everything... but I guess these things are cyclical, and it's only a matter of time before Next & Vercel get dethroned by the next hot young thing. It honestly feels like I'm watching Game of Thrones where every few years all the main characters die and are replaced by new queens. That's the web world for ya. Always has been, I guess.