It wasn't a clean "sweep" for the de-pigmentation genes yet, however. If you look back at the Hofmanova et al chart you can see that there was still variation. Perhaps because they were relatively recently admixed?
As to the Villabruna group, I don't think there's clear evidence yet whether they were originally in southeastern Europe or Anatolia. The Reich group seems to be on the fence about it. Or maybe they were related groups.
The earlier paper on the emergence of blue eyes places it somewhere around the Caucasus. That's probably off, given that the CHG were darker and J1 and J2 probably brought some of the darker alleles to the Levant, but somewhere around Anatolia makes sense to me. That might put it in the people who moved into Europe long before the Neolithic.
For the blue eye snp the whole thing is complicated by the fact that there's no clear "advantage" for selection to operate upon. Of course, there may be something in the background haplotype which is advantageous in certain situations having nothing to do with iris pigmentation per se. There's also sexual selection. You never know what traits are going to "appeal" to certain groups. In some African groups they killed albinos. In other areas perhaps they thought blue eyes were "godly" given it's the color of the sky.
Blonde hair too has no clear "advantage" environmentally in cold climates, otherwise we wouldn't have Solomon Islanders who look like this.
It would be interesting to know if the same snp is involved or if it's congruent evolution. At any rate it brings into doubt in my mind that blonde hair results from a combination of snps.