Here's a different perspective. For context, I don't write much general-purpose software, but I am writing one or two free/open textbooks[1,2] (for statistics/econometrics) and I've thought a lot about how they should be licensed. I think the issues are similar.
My biggest concern in this project is "abandonment." There are lots of fantastic books that have since been abandoned by their authors and are no longer suitable to assign for a class---sometimes the author died, sometimes they moved onto other projects; a few key chapters might still be relevant, but I can't expect students to buy 9 different textbooks and read one or two chapters from each one. There are even completely free textbooks (Bruce Hansen, an extremely well-respected econometrician, has one[3]), lots of teachers have online lecture notes that are quite good, some of the stuff on MIT's OCW are excellent, etc. But I don't know what anyone's plans are for maintaining those resources, what direction they plan to go, whether they will keep them free in the future, and so on.
What I'd really like to do is be able to maintain my own copy of this material so that I can continue to add to it if the original author loses interest, correct typos, etc. For most of these books that's impossible because they aren't licensed for that (even if the book is out of print). But even for Creative Commons books it's not feasible, because they only distribute the pdf. If you look at Hansen's book (link below; it's not CC, but it's representative of the material), you'll see that there's a lot of math in these books, so backing out a human editable version of the text from the binary would be a lot of work.
So, I think that there should be free textbooks and I know that textbook projects are often abandoned (or eventually made non-free) by their authors. So it's pretty clear to me that the best way to have there be free textbooks is to write one (unfortunately, since it's a lot of work, but I do have lecture notes from when I've taught these classes as a starting point) and to use a license that entitles anyone who gets a copy of the book to also have a version they can edit and maintain. The GNU Free Documentation License[4] is the only license I know of that does it (it's the GPL of documentation), so I chose it instead of any of the CC licenses.
The FDL is more burdensome than I need it to be; if someone prints out a chapter of the book and hands it out to his or her class, I don't really need every student to get a printed version of the FDL (I'm fine with the professor just telling them about the FDL), but on balance I think the license makes the project stronger. No one needs to worry that I'll abandon or derail these books, because the license guarantees that they can fork them (no matter how they originally got a copy of the books).
I don't think that people who use CC licenses for educational material are unprincipled. Even though I think that the FDL is a better for projects like mine, it's great that they're making the material widely available (even if I worry that it will eventually be abandoned, the fact that I can redistribute their binaries is very nice); and I definitely don't see why they'd think that I'm greedy for choosing the FDL.
[1] http://www.econometricslibrary.org
[2] https://github.com/EconometricsLibrary
[3] http://www.ssc.wisc.edu/~bhansen/econometrics/
[4] http://www.gnu.org/copyleft/fdl.html