Well, it is giving up freedom, regardless of what it is like.
Keep in mind that there is a strong tradition in political discourse (in the U.S. at least) that using the word "freedom" is a signal that we are in favor of something, and thus, that "giving up freedom" is a negative thing. However, "freedom" is a word with a clear definition, and we can use it without intending the above signal.
So, it seems to me that this policy of Princeton's is a fine idea. It is nonetheless a reduction in the freedom of Princeton's faculty.
In contrast, at my own university the faculty contract explicitly assigns rights in their research to the researcher. The university gets a royalty-free license, but that's it. I have never had occasion to consider this contractual clause in a negative light, but here it is: my university could not enact a policy like Princeton's without violating the contract. (OTOH, my university is not nearly as big of a player as Princeton; at present such a policy would probably not be helpful.)
While that's true, this is a discussion about negotiating in a game-theoretic context. "Giving up freedom" is usually associated with the game of chicken, whereby one party improves their outcomes by "removing their freedom" to select the chicken option. The ggp implies that this was the operating principle, when it in fact is not.
Saying that I have given up the freedom to give other people the copyright to my research, when that was precisely my goal in the first place, is IMHO a clear abuse of language. It's like saying a superbowl team has given up their freedom to lose the playoffs because they decided to practice too much. It's just ridiculous.
Couching this in terms of freedom is the wrong way to think about the issue, methinks - it detracts from the real issue. Freedom in the US is a very politically charged word - as an outsider visiting the US, I could hear a capital F in the word almost every time I heard it in public.
As such I welcome someone using the word correctly for once.
I certainly found when I visited the US for a while that there was nothing I felt I was freer to do legally (tourists don't really get to 'bear arms'), but plenty I couldn't do culturally - ironically mostly about speaking freely in public. I imagine an American visiting Australia would feel similar cultural restrictions, but again, no particular activity they're used to would be forbidden them legally.
I guess it just irritates when you run into those particular folks who define themselves as 'free' and making you out to be 'not free' by comparison.
As others have pointed out, collective bargaining is a particular way of giving up individual freedom.
There are many journals who would be willing to decline to publish the papers of an individual intransigent researcher. There are not many journals who would be willing to ban papers from Princeton entirely.
As an example, Dan Bernstein, who is certainly a well-respected researcher in his field, has individually adopted a policy of dedicating his papers to the public domain so that, like researchers employed by the US Government, he cannot assign copyright in them. The IEEE has consequently decided to reject his papers even when its referees accept them, even though they accept papers from US Government researchers: http://cr.yp.to/writing/ieee.html
This kind of phenomenon, where giving up freedom improves your negotiating position, was extensively studied by Schelling in his theory of negotiation.
It's also very relevant to the debate over Treacherous (or "Trusted") Computing, signed bootloaders, and the like. You might prefer not to have the option of cryptographically certifying to Warner Music that your machine is running an approved operating system on approved hardware. That's because if you have that option, they might not sell you music unless you exercise it, probably giving up many of the rights you have under copyright law; while if you don't have that option, they are faced with a less tempting choice of only selling you a CD.
Advocates of these systems sometimes design systems where you have full freedom to run either signed or unsigned software, claiming that this makes their systems safe from abuse. This overlooks this phenomenon, where having a freedom makes you subject to pressure to exercise it.