I was never thinking of the second option, actually, but thanks to make me think about: I've just heard of some strange histories of brutally rejected papers that have been soon proposed by some other authors, eventually dear and close of the rejecting editor.... Strange, is it?
In chemistry, the main reason people don't use a preprint server is that some of the major publishers -- most notably ACS -- don't allow it for papers appearing in their journals. It's a difference in culture, really: physicists all tend to use preprint servers, so the publishers are kind of forced to go along. In chemistry, the publishers have managed to clamp down before a critical mass started doing it.
In physics and mathematics the use of arXiv is very common, see .g. the first link below. Biology has fairly recently got biorxiv -- a counterpart of arXiv.