Who forges the partnership?

Also, the census data issue for the U.S. is a very real one (with all the rambot articles, as well as many uses for the census bureau's geodata). Issues with census data is something OSM has to deal with too, having imported a lot of their geodata into OSM. As step #1, I am working to setup a spatial database with some census geodata, and work with it in similar way as using osm data to make maps.

I'm not sure what role Wikimedia can play, but more support for the OSM-Wikimedia infrastructure work (both hardware and developers) is probably needed. Since we are getting into content issues, the foundation does not necessarily deal with content, so perhaps some of this should be handled through a chapter? (we don't have a U.S. chapter, just the NYC chapter which i doubt is interested at this point, nor has the capacity)