For some people, the process of research is very organized and methodical. For me research is something creative and instinctive, slightly chaotic and usually a bit messy. I am not saying that one way is better. I know that there are distinct disadvantages to my method, such as the possibility of finding a really good source through some odd combination of luck and insight - only to lose it, never to find it again because of being disorganized. So I've been looking at better ways to be organized - but not in a way which slows me down and causes me to lose my momentum. For me, research is the closest that I'll ever get to hunting. One of the things I like about it is the chase, which is why I hate things which slow me down
I have set up a separate Gmail account for my work and use it as a cache of my research. As I come across articles and sources which seem promising, what I do is compose a new email, paste in the full text and citation/URL of the article and then save it as a draft. It's a very quick process, and so I'm able to move on. The nice thing is that because Gmail is searchable, everything in this research cache is searchable - making it easy to return to the documents to dig more deeply and hone in on what is useful. It doesn't work so well with pdfs, but I can still upload the pdf and store the citation and/or URL. The other nice thing is that I can add to and search this cache whether I'm in my office or at the shared computer on our reference desk.
I am quite confident that because this research cache is personal - and can only be used by me - and because it's do with my work which is for academic purposes, that in most cases, it would be all right. To be absolutely sure, I would need to check all of the license agreements of all of the database which I am likely to use.