Independent of any equity arguments in any direction, I believe contributor agreements are usually a real mistake, and should be avoided whenever possible.
1) getting them "right" is very hard. I would not/could not sign the Fedora agreement for years due to language that made it unclear exactly what I might be signing away rights to. At the time, I was on the X.org board, and that agreement was written in a way at that time that made it unclear if I might be speaking on behalf of the parent upstream organization, X.org (as opposed to whatever Fedora specific work I might be doing for which I have little problems giving RH rights to).
2) if you work for a corporation, you will usually have to have the agreements reviewed by legal counsel; and it effectively doubles the amount of legal effort to get something "over the wall" to benefit anyone. The inertial barrier is already high enough. And with the problems I covered in 1), you can get into dread "lawyer wait", where no one wants to take any responsibility for their actions, or because the agreement is just hard to understand and therefore review is not timely.
If you are just publishing code under a "standard" major open source license, then legal departments generally don't have to understand the licenses again and again; they internalize the specific licenses.
So contributor agreements between contributors (in a company) to a project like Ubuntu means that we go from a order N problem (of licenses) to add contributor agreements to each project that requires them.
- Jim