That's a great idea! I've heard of researchers doing this kind of analysis to detect "hidden gems," i.e., papers that influenced papers that ended up becoming the big cited paper. The tough part, of course, is that most journals are keen on keeping all of their data behind company walls. Perhaps PLoS would be an exception, though...
I've used such a tool at university but I've now found out it's a subscription service. You could get graphs of papers that were cited and could use it to find papers that were roots of a certain discipline. Though it wasn't as advanced a graph analysis as you propose it was pretty helpful.
And you're right about journals sitting on that citation data. The service I used was
