It's interesting how this "little" detail always seems to get overlooked by the XP zealots. A great book which I think everyone should read before jumping onto the XP/TDD/Scrum/Agile/etc bandwagon is called "Extreme Programming Refactored: The Case Against XP". It highlights the failed C3 project and offers some compelling arguments.
Don't conflate having a project canceled with having it fail. Successful projects get canceled for a variety of reasons, and some failed projects never get canceled.
