How far can you get from a McDonald's in the continental US?
weathersealed.com
weathersealed.com
I ran across aggdata some time ago while looking for a list that they happened to provide for free. They provide niche data that they scraped from the internet. Their data is cheap enough that it's not worth trying to scrape the data yourself. I think the business model can be duplicated easily and there is plenty of room on the long tail.
I once scraped the data for Starbucks in Manhattan and it was a bit of a pain: http://gawker.com/011483/gentrification-map (2003? I feel old.)
Comedian/Filmmaker Mark Malkoff attempts to visit all 171 Starbucks stores in Manhattan in one day.
Probably the South Pole.
There is also a really nice Arby's there.
src: http://www.reddit.com/r/geek/comments/9nbyc/on_a_recent_road...
http://maps.google.com/maps?f=q&source=s_q&hl=en&...
It's extremely desolate: it's very cold in the winter, hot in the summer, and dry year round, so nothing much grows out there.
But it's a fascinating place to explore if you like big empty spaces (or at least like them once in a while - I wouldn't want to live there).
Oh man, you have no idea; some really good shit grows in that part of the country.
(I guess coming from Australia I might have different standards of "big" and "isolated" though.)
edit: http://mcdonalds.com.au/find-us/restaurant shows us that my off-the-cuff guess was largely correct. Zooming in on their handy map even gives us a nice rough set of directions: go along the Gary Highway until you reach the turnoff for the David Carnegie Road, and you will be at the least McColonised place in Australia.
--- Between the tiny Dakotan hamlets of Meadow and Glad Valley lies the McFarthest Spot: 107 miles distant from the nearest McDonald’s, as the crow flies, and 145 miles by car! ---
http://www.nationmaster.com/graph/foo_mcd_res_percap-food-mc...
For free, you can use the Many Eyes project to do this kind of visualization as long as you can transform your data source into the required format.
There was also a neat proof of concept that someone from the OpenLayers community did a while back:
http://crschmidt.net/mapping/choropleth.html
I've used that code in my Mozilla Metrics to generate some nice choropleth maps.
Finally, there is the code that I worked on as part of the Mozilla downloadstats visualization. It reads a JSON feed of locations and puts markers on a Canvas map. http://www.mozilla.com/en-US/firefox/stats/
(Although it is the only state to not have one in the capital city, also the only state to not have one in its largest city)