> It would be nice to see an effort to migrate some of the kernel core, but I can't see Rust gaining widespread acceptance in the kernel development community any time soon.
I don't even think that a migration effort from C -> Rust for the Linux Kernel is even feasible or even worth it, given the scale of the project. At this point, you might as well start from scratch.
Google is already experimenting with Rust for OS development with their new Fuchsia operating system [0], which has some drivers written in Rust with a capability based security model.