Interesting analysis. Autosomal DNA is particularly difficult to use to determine migration patterns, so this is a commendable effort. It seems to leave an odd contradiction... while Y-DNA suggests that Eastern Europeans are more closely related to the initial proto-Indo-Europeans than Western Europeans are (in that R1a extends into Asia more significantly than R1b), autosomal DNA does not support that.
I will say that I don't understand the exercise of trying to find "any population left today with such proportion." Surely we don't expect PIE people to still be around, and all the possible areas of the world in which PIE could have originated have experienced subsequent migrations, and probably genetic shifts, as well.