stubgen should try and figure out all the types

In most cases, mypy should be able to figure out the bulk of missing type annotations.

Currently `stubgen` does semanal, but it could use 'checker' as well.

Start by locating the stubgen entry point and the checker it could use, then determine how inferred types should be transferred into generated stubs. Done means stubgen can infer the bulk of missing annotations through the checker, with behavior validated by the project's relevant tests.

Written by the indexing model from the issue text.