My idea of programming languages is that they are better when they are designed with their use case. And in the case of Rust, it is to make a web browser engine. I don't know what the current situation is with Servo but I think that it would be a good thing to keep a privileged relationship between the Servo team (or what is left of it) and the Rust foundation.
Rust is not for every project, in fact it is terrible for many projects, but if your project has the same needs as for a web browser engine, then it is great. By making the foundation too independent, we may end up with a monster (no need for another C++), a useless "jack of all trades, master of none", or simply lacking features that are essential for making a good web browser because it is not good enough for some committee.
Languages need a strong direction, with specific goals in mind. Otherwise, they just wind up being a little bit of everything and not really great at anything (or wind up with a dozen duplicate implementations of the same idea).
I don't think Rust should be limited to just a browser engine language though, multiple companies are already looking at it to 1:1 replace some C++ for its security improvements without sacrificing significant runtime performance.