I've been trying other populations from Eurogenes 15 with this model, and here are a couple more:
| R-Squared | Anglo-Saxon | Celt | Bell Beaker | Intercept |
Danish | 0.995665 | 0.57223 | -0.02596 | 0.452749 | 0.006023 |
North_Dutch | 0.995606 | 0.672777 | -0.1185 | 0.450742 | -0.03299 |
These two seem to fit the model very well, and come out as a mixture of the Hinxton Anglo-Saxons and the Bell Beakers, with a bit less of the Bell Beakers. Maybe they correspond to the Angle, Frisian, Jute and Danish Viking invasions of England. The West Scottish and Irish come out as the opposite, with the same components but the large majority being the Hinxton Celts.
Now here are the Germans:
| R-Squared | Anglo-Saxon | Celt | Bell Beaker | Intercept |
North_German | 0.991923 | -0.19132 | 0.663382 | 0.505257 | 0.151681 |
West_German | 0.962378 | 1.570118 | -1.66584 | 0.991526 | 0.692574 |
East_German | 0.965777 | 0.686971 | -1.81043 | 2.148594 | -0.1685 |
Looking at the North Germans, I thought maybe the Saxons brought over a Hinxton Celtic-Bell Beaker mix. (The South Dutch and French were also high in Bell Beakers) But then, the West and East Germans had more of the Hinxton Anglo-Saxon component and less of the Hinxton Celts -- although their models didn't fit as well.
Now for the Scandinavians:
| R-Squared | Anglo-Saxon | Celt | Bell Beaker | Intercept |
West_Norwegian | 0.990316 | 2.251309 | -1.97525 | 0.757186 | -0.22386 |
Norwegian | 0.986636 | 2.07966 | -2.0498 | 0.991973 | -0.14615 |
Icelandic | 0.993988 | 0.742832 | -0.2822 | 0.58937 | -0.33384 |
Orcadian | 0.998015 | 0.547285 | 0.498763 | -0.06765 | 0.144096 |
Swedish | 0.970309 | 1.96859 | -1.97326 | 1.019808 | -0.10206 |
The Norwegians, Icelanders and Swedes had lots of the Hinxton Anglo-Saxons plus some Bell Beakers, and were low in Hinxton Celts. However, the intercepts were higher, thus suggesting additional components and confounding the results more. The Orcadians had a more even blend of the two Hinxton groups and had less of the Bell Beakers, which made sense to me, seeing how different from the rest of Britain they come out as on other genetic population measures. I don't see much evidence against them being largely a mixture of the West Norwegian and West Scottish samples. All the other populations I tried had lower R-squared values.