Isn't that what Rust is attempting with Cranelift ? They had making Cranelift backend "production-ready" for development use as a goal for 2026. I am guessing it will be a few years beyond that before it is made available for general production-ready use.

I think Zig might possibly beat Rust's timeline here for a "No C/C++" toolchain. That is if its lead doesn't burn himself out.