When working on #192 I saw there's a chance to automate the processing of the datasets related to USA (counties and zip codes at least) following a similar approach as we did for #188 with a Makefile. That way we can declare the outputs and define the commands needed to produce them.
