making the case for the JIT as a supported, non-experimental part of CPython

describe a JIT infrastructure that can support multiple implementation strategies

Mainline branch games. A "proof of worth" scenario is a terrible situation.

Why not keep the JIT and let it evolve alongside the new framework?