> Second, figure out how many clinical/surgical encounters a physician in training needs to truly be competent when they complete their training.

This seems like the sort of thing that should be widely studied, but I can't find the terms to get nice Google Scholar results. In my field we do this sort of assessment for the users of (fairly low-risk) medical devices. It requires a team and is quite work-intensive to assess, but the stakes for insurance providers are high enough that this must already be known....