Within my field, it looks like science twitter is useful, but only up to a limit. The primary use derived from it is a list I found myself added to. It actually

Within my field, it looks like science twitter is useful, but only up to a limit. The primary use derived from it is a list I found myself added to. It actually contained people who were only peripherally related to the research I do, but meant that there was great diversity in the papers that were promoted in the feed - not all relevant, but at least interesting. A second use is a signalling factor that you are active in the field - really an exercise in building brand. Finally, there are probably two people who actively engage in discussion on Twitter (from my field). This doesn't happen more often because scientific review is a subtle beast, and doing it in the open with 280 chars, as quickly as possible on a phone isn't conducive to constructive feedback. In fact, I've got a number of draft tweets initiating discussions that I didn't end up sending because it just felt plain rude to review in that manner!