This only makes the $20K to Mercurial seem weirder, don't you think? It underscores my point that it's hard for large organizations to keep their official positions on things straight.
I think this depends on what you mean by "the less popular things that need to get done." In my experience, things that were less popular at other places I've worked, are often tackled with enthusiasm at GitHub. When the company culture leads you to feel passionately about building something lasting, going back to clean up old rotting code or fix tests that were not written well or debug obscure issues become things to be proud of.
Also, I believe that often the less popular work corresponds to things that are subtly judged to be less valuable work. At GitHub, I see people praised for tackling the types of things that have been less popular at my other jobs, and everyone here, even the most prominent GitHubbers, participate in most every aspect of maintaining the product. Internally, the message is pretty clear - doing the things that "need to get done" is a huge part of what makes everyone else here proud of you and is valued highly.
This is not to say that people don't tire of working on particular areas, but that's a somewhat different issue and probably a large enough topic for a writeup of its own.
