We launched the crawler a while ago now and the results are impressive and even a little surprising in some respects, while in others they're not. Basically we're happy we found a few thousand pages with geotags on the first crawl though the longer we've let it go the lower the ratio gets. Totally expected because we seeded it with pages we knew were geotagged in the first place.  To get more pages, we've started using the mechanical turk. We're hoping to get pages that aren't otherwise tagged added by real people. Once we launch this is obviously going to be a function of the site, something we want any user to be able to do. In the meantime I'm tweaking our turk jobs and starting to get some good data back, though not as much as we'd like - we'll probably have to up the cost.  I'll let you know how it turns out, in the meantime if anyone has any experience with turk, any hints or thoughts - let me know.