If you need to use a C level inline keyword then you can always do it in Cython (which explicitly supports it, although only for functions that you’re calling from other Cython code).
To me this sounds like the most sensible solution - if you want to force it inlining then you’re probably better doing it with one of the various Python accelerator tools available.
Obviously if Python wants to implement/improve bytecode inlining to speed itself up then that’s great, but probably not something the user wants to be thinking about.