Actually, I would really like to see a navigable map of science. What web of science has is weak (you can check two citations back and forward) but it really do

Actually, I would really like to see a navigable map of science. What web of science has is weak (you can check two citations back and forward) but it really doesn't let you check out nodes or unexpectedly valuable but potentially under-cited papers in your field. (Think a paper that several big papers cite but no one else seems to.) It would be nice to look at a field and be able to figure out what links together and what areas are connected well and which are poorly. (A possible indicator of people studying an artifact? Or perhaps an area missing out on a possibly useful technique? Would over-linking mean a trendy field? What would a broadly connected field mean? What would an incestuously connected one mean?) This would also be a good way to find the 'best' reviews that everyone cites. Add in a way to sort by year, field, and to rapidly zoom in and out on papers and topics and you'd have a fantastically powerful way to visualize science. It would be even better if you could 'play the tape back' and show what the network looks like year by year. Do fields follow a stereotyped network behavior before exploding? Or does each explosion look different? What does the citation map of an explosion look like? What does a field that is about to 'die' look like?