> If Torvalds can do it, why not?

I genuinely doubt Torvalds (even a young version of him) could build a new competing OS in modern times, and have it succeed.

Linux succeeded in a much, much simpler world. What needs to succeed now is something that compares not to the initial Linux, but to the Linux 30 years later, with mountains of effort piled up on top. That is no longer something that can be achieved by any single person, no matter how motivated.

There are a few projects for Rust only OSes. Redox aims at the desktop, Hubris aims at microcontrollers, to name two I'm aware of.

But there's a huge difference between "let's make it easier for driver writers to write correct drivers by using Rust in the OS we all already use" and "let's create a whole OS and convince driver writers to write drivers for it and then users to use it".

Linux kernel devs want to work on Linux, not on other OSes.

Companies depend on Linux and want to improve their products, not rebuild a massive codebase from scratch.

Reaching parity with Linux's immense suite of device drivers is perhaps the single biggest hurdle.

That's one of the biggest reasons why alternative kernels either remain fringe or fail.

Initiatives such as NetBSD's Rump kernel but for Linux may provide a bridge to Linux's drivers but it's brittle. There's already LKL - the Linux Kernel Library project with similar aims but not much traction.