Dienekes has a post on a new AJHG paper by Marc Bauchet, Mark Shriver and others on European substructure. Dienekes covers it pretty well, including the picture of the Structure output of K=2 to K=6.
This study looks at 297 people from Europe with 10,000 SNPs
Like this previous study by Seldin et al. , the Finns are more in the fuzzy area of the Europe-Asia gradient (or just off on their own).
Surprisingly with 10,000 SNPs, it seems to be hard to resolve much substructure in Europe - the yellow part of the Structure picture is not broken up much, except for the Finns and the Southeast Europeans. The sample sizes from each country are quite small (about 10 per country).
They do provide a list of SNPs that are especially informative for distinguishing northern from southern Europeans.