I’ve been keeping my eye on the Ancestry update for a while. If you didn’t think that the new panel could be skewed, you were wrong. Example:

  Albania
128
  Anatolia & the Caucasus
2,000
  Arabian Peninsula
1,260
  Ashkenazi Jews in Eastern Europe & Russia
2,054
  Ashkenazi Jews in Central & Southeastern Europe
2,291
  Azores
1,501
  Basque
253
  Bengal
1,146
  Benin & Togo
1,138
  Brittany, France
85
  Burusho
17

Clearly, 1:1 ratios with minimum samples of 400 (though ideally 1000) in all groups are not present. That alone skews the results.