I suspect this is not a problem specific to OS development. I've experienced the same bias towards new features from within a dev team. It seemed to be down to a percieved need to keep up with or ahead of the competition in terms of feature complement. Plus, if a problem had gone unfixed for quite a while it was often because it would require more time and resource (which could otherwise be working on new functionality) than it was deemed to be worth. It would be interesting to hear if/how other people balance focus on bugs v. shiny.