RPython is a framework for building trace-based JITs. Its selling point is that it automatically produces a JIT from a high-level specification of your language (basically, you would implement an interpreter for JS in RPython and then the framework would spit out a JIT)

In JavaScript land there is less space for high level frameworks because browser vendors are willing to throw thousands of man-hours into JIT development. They are willing to go through this trouble to avoid the runtime overhead of meta tracing.

Additionally, JS JITs have moved towards a more method based than trace based approach over the years, for whatever reason.