The most interesting thing about this article to me is where the author got the data -- http://www.aggdata.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've actually been there. It was a surreal feeling, one that is hard to explain in words. There was something about that place that made me feel alive, free from constraints, for some reason I could think clearer, and hunt for the answers.
I was in Death Valley a few weeks ago and I was surprised to find it wasn't all that big, or all that isolated. Heck, most of it is within 100 miles of Las Vegas (and more importantly, it's all within about 70 miles of the McDonald's in Pahrump NV on one side and Lone Pine CA on the other).
(I guess coming from Australia I might have different standards of "big" and "isolated" though.)
That would probably be around the middle of Western Australia, since there's a Macka's in Alice Springs that throws out the absolute middle of the country as a candidate.
Also, if you look at a road map, that spot in the middle of WA is the furthest from major ("interstate") highways.
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.
The answer, for those too lazy to read the article:
---
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!
---
If you like this kind of stuff, then OpenStreetMap is for you. Because it is free and open you can use the data to create such visualisations easily. Granted of course, that someone (how about you) added the information you want to look at.
Does anybody know how that map was generated? What software can easily assemble geolocated data into a nice glowing map? I assume you could do it manually via processing or similar, but there's gotta be something pre-assembled.
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/
No its not in the CONUS, but many moons ago I was at Fort Greely, Alaska. When I was done with my tour and if conversation turned to the topic of my past travels, people would ask the whereabouts of Ft Greely. I say it was 100 miles south of Fairbanks. I usually get a small nod. When I then followed up with "and 90 miles from the nearest McDonalds" their eyes usually got bigger.
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.