It places many restrictions on both derivative work and on the source (not just source code) work. Lets list some of them (we can just stick with v3 for now):
1) If you want to give a copy of any part of a work (say the binary executable), you MUST also give the source code and additional resources (via one of many methods).
2) If you convey (see v3 definition of convey) the work then you agree that any patent you own is licensed royalty free to any recipient of the work (at least while it relates to this work or a derivative of this work).
3) Unless every copy of a work that you convey under the GPL also includes the full source and other required items, you are required to keep conveyable copies of that source for the lifetime of the license (or until you or your company ceases as a legal being). (See GPLv3 6 d)
4) You must include or maintain "instillation instructions" for the work. If you have ever maintained a large software project you will understand that this is not by any means a simple task.
5) You are required to use proper attribution to previous authors as noted in the license.
6) Section 7 also adds in multiple additional restrictions that are optional to each copyright holder to their works. These are restrictions that are specifically listed as being acceptable as part of the GPL.
Each of these items is by definition a restrictions on your "rights". I am NOT stating that this is bad. It is simply insane to try and prove your argument by stating things that are unarguably false. It hurts the case of the GPL and undermines the support of the OSS community (regardless of what camp you are in).
The GPL is a useful tool exactly because it DOES place specific restrictions on what you can do with a work and its derivations. Your argument should not be that it does not act in this way, it should be that the restrictions are reasonable, useful and outweigh any negative factors.
This fixes a fatal flaw in BSD: in BSD, anyone can place copyright restrictions back.
Again this is false. If I write a section of code and distribute it with a BSD license, no restriction can ever be placed on MY code by anyone else (I can still release it again with other conditions). Others can take my code and modify it and then place restrictions on what THEY wrote. But no restrictions can ever be placed on what *I* wrote. This is similar to me walking through the park and picking up litter, I don't care what other people use the park for, I am merely giving myself and everyone else a gift.
Again, I am sure you don't like this method, but that has nothing to do with statements of fact. In this case you weaken your argument (and that of the OSS community) by making statements that are simply and provably untrue.