I believe that Code Development is a form of Research, and that Open Source is Peer-Reviewed Publication. If those two statements were widely accepted, teaching Open Source would be a lot more paletable to many university professors. We must advance that position while accommodating the transition: in comments on Greg's posting, Dr. Jef Spaleta noted the need for a peer-reviewed journal on Open Source. I believe this could be a very positive interim step. (Jef, are you going to make it happen?)
In the shorter term, we need to ensure we understand one another when talking about teaching Open Source, and its become painfully obvious that the participants in many conversations are not even on the same page in terms of what it means to teach Open Source. In an attempt to help sort this out, I've created a rough sketch of a taxonomy of Open Source education. Please take a look and join the discussion.
These are my first two books: X Power Tools, a thorough guide to the X Window System (O'Reilly, ISBN 9780596101954) and Fedora Linux: A Complete Guide to Red Hat's Community Distro, a practical hands-on book on Fedora (O'Reilly, ISBN 9780596526825).
Fedora Linux is also available for online reading through Safari and in downloadable PDF format from oreilly.com