Unless I'm not seeing some other visualization of the data, I don't really see how correlating companies to physical locations helps a whole lot, and was not what I expected to be the goal when I started reading. The pain point that needs to be addressed is "too big to fail", but mapping to physical location does nothing to help that.

A more conceptual map showing, for instance, the players in the mortgage space, their relative size, and how they generally interact would actually be a step towards solving the problem. Now how to gather information like THAT is a whole other question.

As a concrete example: Most of the major home insurers have offices in Florida, but don't actually write policies. Who is actually writing these policies? If we have everything tracing back to a single policy writer, then we have a major problem.