I am also someone who just wants to get shit done. I've never had an issue with php, c#, ruby, crystal, go, rust, haskell, erlang etc's compilers/interpreters in terms of letting me get what I want done. Yeah, I haven't actually forked or patched those compilers but neither have they been hostile to their users.
The reason I would use literally any of those except Elm is that the maintainer(s) of Elm act like they're Steve Jobs. How can I "get things done" if I'm being told "oh no, you're not allowed to do <standard-thing-X>, we've decided that only us on the Elm team can do that". With other languages, I'm not worried about a version bump pulling the rug out under me, effectively trapping me in the ecosystem unless I want to rewrite everything in a different language. That is also directly the opposite of being productive and getting things done.
There is nothing about an autocratic "you're holding the compiler wrong" governance model that makes any part of a language itself more efficient or fun. Plenty of other languages are as (or more) efficient and fun as Elm, but without the patronizing and condescending attitude of "you're not as smart as us, so do as you're told, and no you can't discuss this with us". There is no way I could justify the risk of adopting elm professionally given the entire rest of the programming language world... doesn't do this. Let me be the one who decides to shoot my own foot off; even Rust (a language pitched on memory safety) allows this with unsafe, etc.
If one wants to control everything exactly and go against almost every other language and most F/OSS projects, then just... make proprietary/source-available software. F/OSS is a give-and-take, and nearly all of us have taken far more than we've given.
And I don't know what John Carmack's "proverbial engineer" is, but I think just about any engineer would prefer to be in control of their tooling and not at the whims of maintainers.