the studies these graphs are based on look like they might lend support to the notion of quadra values and integral national types. my guess is that correlating the corners of the graphs to quadras will be a bit of a can of worms though.
for more information:
http://www.overcomingbias.com/2009/1...ed-values.html
http://www.overcomingbias.com/2012/0...2d-values.html