I'll try to summarize some of the issues.
First, there is the concept of the stapled thesis that is common in some fields/places. The thesis consists of a set of published papers (published) and, perhaps some unifying commentary. The thesis itself is (largely) just a collection.
Publishers usually require that you give them copyright to things of yours that they publish, and will give you back a license for some (but not all) uses. That license, these days, will almost always give you rights to include the work in a "stapled" thesis (I don't know of exceptions, actually, but must admit the possibility). Generally, though, a stapled thesis probably isn't considered by the publisher to be a "republication" of the already published work.
Somewhat orthogonal to that is the notion of self plagiarism. This is a form that has only some of the characteristics of (ordinary) plagiarism, though without misattribution of the person who originated the ideas. Self plagiarism is considered a violation because if you, for example, copy-paste your own earlier work without citation, you break the chain of evidence for discoveries. The early work contains context that a researcher wants to see along with the current work. You avoid self plagiarism by quoting and citing the early work, permitting that chain of context to be followed.
A third issue you raise opens the problem that if you put unpublished work in a thesis (your unpublished manuscripts) and if that is considered publishing then some publishers won't accept it. That is becoming somewhat less of a problem these days with preprint servers (and is not much of a problem at all in some fields that welcome preprints), but you need to be sensitive to that. But, if you have unpublished work, you can do what you like with it, copy-paste, whatever. It just might limit what you can do afterwards.
Generally speaking it is better (and a better habit) to keep things clean. If you copy, do so judiciously and with citation. Make sure that your license (from published work) permits what you want to do. Ask if you aren't sure. If you cite, you avoid hints of plagiarism.
If you want to write a non-stapled type of thesis using early work the best path is to treat the work as you would the work of anyone else. Quote from it, but not too extensively, just as you would from the work of others, and cite what you use. Then, if the thesis contains new results and extensions to old results, it might just be publishable as a new work. I'll also note that the license from a publisher will probably give you rights to copy somewhat more extensively than you could from other works, but likely not unlimited copying.
A major issue for you, however, is what your university expects in a thesis. If they accept a stapled thesis, you are fine. Write up the "bridge/summary" commentary if needed and staple it together. If they accept a new summary of old work, you are probably fine, but keep it clean - quote and cite. But if they expect an entirely new work (unlikely under the circumstances) then you need to provide that.
It would, however, seem odd to me if the university permitted unlimited copy-paste without quote or citation. That is just poor scholarship.
I hope this is enough commentary to guide you.