See Hammond et.al.(2004) for a nice review of the current state of RSS and it’s existing role in scientific publishing (in dlib).