He has just done a followup on his own site, about some preceding Chinese and Western exercises in the same spirit. Very much worth reading, here, along with this post on a related theme.
Update: To be clear about it, any suggestion from the discussions above that Patrick Chovanec's map was in some way "unoriginal" is entirely unwarranted, from my point of view. The concept that big, monolithic "China" is better understood as a variety of diverse sub-units is a well-established, even obvious one. The plus of Chovanec's presentation is the execution, which makes the point interesting and accessible for people in a new way. And as he points out on his own site, in his initial (very long!) submission to the Atlantic, he catalogued a variety of previous efforts in this same direction. That history would have fit well into a long, print version of his analysis, but not so well into the kind of interactive online feature we have presented. So, congratulations to him for making an important concept interesting and vivid for readers.