I've only found time to do this with CherryPy really, mostly cause every time someone asks for help they say if something is missing in the documentation you could always read the source which is also commented and readable. It really is approachable and not too big. I never have time to go through the source of every single library but you raise a good time to do so: stack traces. Although when doing front-end JS code it's a bit less approachable if your files are already minified (client hands source like this) and the error becomes a lot less useful.
