Oh, that's a bold move. I doubt this would happen on ClojureScript, in fact, I see the opposite sometimes, some people in the community complain about how Java and JS ruins Clojure/Script and makes everything worse, and why isn't there a native Clojure, where you'd never have to deal with the host. And the core stance tends to be that being hosted is the whole point from the beginning.
I can see the frustration with Elm though, especially that it was enacted as part of a version upgrade. If it was from the start, people would know what to expect, but this seems like a huge backward breaking change. I'd be annoyed as well.
In Clojure/Script, a lot of the tensions are more about what should be a community provided library, and what should be lifted as officially included in core. I see the core team tends to favour most language extensions to be kept as a library. And often when people want changes to the language to be made, the answer is that Clojure/Script is designed to allow user level changes, so whatever you don't like you can change for yourself. Some people still arn't happy about that, they want their ideas to become the standard.
A lot of that stems from it being a Lisp. There's very little of Clojure/Script that actually necessitates changes to the compiler itself. Even the core functionality, most of it is implemented on the basic primitives the compiler gives, so you can happily change and extend almost anything if you disagree with the core team's choices.
Back to Elm, I know personally, I've never been into a language restricted to only one use case, the web. It seems Elm has always wanted to be more of a framework for web development. A very opinionated one. This seems to be a move even more on that direction. Like if Ruby didn't exist, but only Ruby On Rails did. Correct me if I'm wrong here. It could lead to something nice in the long run, but like any framework, the trade off is that when it doesn't have the feature you need, you're stuck.