i disagree. this is a great project. the project demonstrates the ability to quickly cull information from different spaces and combine that information via some algorithm for presentation.

what i see are two separate achievements here. collection and analysis. the collection looks good and they have implemented some algorithm. the efficacy of their algorithm may leave something to be desired in terms of accuracy, etc. nevertheless, the algorithm may be tuned/replaced/compared going forward.

i wont even go into the many many applications that something like this can have.

I came here to say exactly this...

The actual search probably takes only several microseconds and was probably trivial to write compared the to the pretty visualization.

Yeah, much too much "art." I wonder what it would be like with better backend intelligence-gathering software, like Maltego http://www.paterva.com/web4/index.php/client# ?