"I’m dead serious—and I can afford it about as well as you’d think I can." See his blog.
Similarly, you'd better expect that the professor will go find another research assistant to work with.
I'm a professor, though not at a research institution. Here's what I would do if I were and hiring research assistants as bitchy as the poster...
"Want to be my research assistant? Then sign this. Yes, your work becomes my property." "Oh, don't like that? Why don't you go find another professor who is hemorrhaging grant money."
Seriously, why would I need, let alone want, to deal with some FNG with very little experience,
full of himself, fantasizing that he's got the next killer break-through rattling around his excuse for a brain pan?...
Well, I'm a professor at a research university, where most Ph.D. students are RAs (except while they TA or have outside fellowships). Several of my Ph.D. students have gone on to be professors at top-ranked universities, so I'm probably at least an okay advisor. So let me tell you that advising Ph.D. students is all about respecting them and their ideas and opinions. It's also about trying to instill good taste and values in students. I am shocked to see someone who claims to be professor have so much contempt for his or her students.
As for licensing software, I always explain to my students that they should make their projects free software to have the most impact. I discuss the options with my students, but generally let the lead student on a project select the particular license, ideally with rough consensus of all involved. So yes, even though the university might own their work, my students are free to continue using it and building on it in perpetuity.
It would be wrong for me to confiscate students' intellectual property--particularly if I tried to make them sign something saying their work belonged to me, as opposed to the university. Moreover, it would be setting a terrible example and instilling bad values in students. Finally, it would probably be illegal, because the university has policies in place to prevent the abuse of students.
Yes they can take away all your rights to the software, but no you shouldn't allow them to do it.
First, I've been a grad student at one university and a professor at another, and I've always avoided signing these agreements. It turns out that if you just avoid signing them and aren't too confrontational about it, you can easily slip through the cracks.
Second, you should talk to your professors and see if they will allow you to develop software publicly under some irrevocable license like the GPL or BSD. With revision control software like git, it's pretty easy just to throw the repository on your home page and make everything you do available to the world (including yourself) on a royalty-free basis. Import some GPL-ed third-party code into your project for extra protection.
Finally, sometimes professors do try to exploit grad students for the purposes of launching their startup companies, etc. If you feel that you are going to be in a position where your research is compromised (for instance because your results are no longer reproducible by the community), then you should find another research group to work with!