First of all, congratulations for the great job. I love JabRef, particularly for its clean interface that at the same time incorporates many useful functions, but perhaps more important, because JabRef don't force users to store files in a particular place or way.
Now, one feature that I think would make JabRef even better is the ability to extract comments and highlighted text from a pdf file. Docear has implemented this function to work with a number of pdf-readers. In JabRef these annotations could be extracted in separated (or a single) fields, with (but preferable without, to keep it simple) the page where the annotation came from.
Regards
Modern request at https://github.com/JabRef/jabref/issues/662, therefore I close it here.