Would really like to compare Theopetra and Sarakenos to modern populations. And also o Iron Age Paeonians, Illyrians and Logas. I wonder if the 'Balkans' encirclement includes mainland Greece. Seems to me that Sarakenos is very similar to Cretans, while Theopetra might be very similar to mainland Greece today.
So far Sarakenos, Theopetra, Logas 01 and 02, Crete_Armenoi and the Marathon sample seem to be closer to modern Greeks than the four Mycenaeans of the Lazaridis study. And out of those four Mycenaens, two were from Crete. I leave out all the other specimens, because they are either from Sicily or Spain. Who knows who they may have intermixed with. That makes me think whether the earlier image we had developed for Mycenaens is correct. We only have two out of a total of zeven Ancient Greeks from the mainland with a low rate of steppe vs Minoan (1:10). They other five are more like 1:5, or even 1:4.
You say that in Lazaridis' Minoan/Mycenaean paper, 2/4 Mycenaeans are from Crete? This is false. 2 were from Galatas in Argolis, 1 was from Pylos in Messenia, and 1 was from Salamis in Attica.
Additionally, you seem to ignore all the new Mycenaean samples from the recent Southern Arc paper, which added an additional 11 new Mycenaean samples from mainland Greece, and also 2 archaic Greek samples. Of the Mycenaeans, 6 were from Pylos, 3 were from Attica and 2 were from Kastrouli in southern Phocis.
While there is variability in the amount of Steppe, their average is not so different from the Mycenaean average that we had previously.
To make things more interesting, when averaging all 15 Mycenaean samples together, their average is almost identical to the average of the genetic cluster 480BCE_1 from Himera, which the paper identified as definite Greek soldiers. They are also very similar to the two Phocaean colonists from Empuries in Spain, and the Thracians from southern Thrace.
[TABLE="class: distances"]
[TR]
[TH]Distance to:[/TH]
[TH]GRC_Mycenaean_(n=15)[/TH]
[/TR]
[TR]
[TD]0.01049552[/TD]
[TD]ITA_Sicily_Himera_480BCE_1[/TD]
[/TR]
[TR]
[TD]0.02023135[/TD]
[TD]Iberia_Northeast_Empuries2[/TD]
[/TR]
[TR]
[TD]0.02360508[/TD]
[TD]BGR_KapitanAndreevo_IA[/TD]
[/TR]
[TR]
[TD]0.03057226[/TD]
[TD]BGR_Anc[/TD]
[/TR]
[TR]
[TD]0.03125336[/TD]
[TD]GRC_Koufonisi_Cycladic_EBA[/TD]
[/TR]
[TR]
[TD]0.03189316[/TD]
[TD]BGR_IA[/TD]
[/TR]
[TR]
[TD]0.03294215[/TD]
[TD]ITA_Prenestini_tribe_IA_o[/TD]
[/TR]
[TR]
[TD]0.03420535[/TD]
[TD]GRC_Kastrouli_Anc[/TD]
[/TR]
[TR]
[TD]0.03561312[/TD]
[TD]ITA_Tarquinia_Imperial[/TD]
[/TR]
[TR]
[TD]0.03846649[/TD]
[TD]TUR_Aegean_Izmir_Yassitepe_MBA[/TD]
[/TR]
[/TABLE]
I'm also not sure why you think either the Empuries or the Himeran 480BCE_1 cluster show any signs of having non-Greek admixture. A shift towards Spain and/or Sicily is usually pretty apparent, especially because Greeks of this time period lacked HG admixture which is more abundant in western Europe. A few of the 409BCE Himera samples show an obvious shift towards the Sicani for this reason, as highlighted in the paper itself.
Now, I do agree that the Theopetra samples are certainly interesting, but don't forget that 1. they predate the Mycenaean culture and 2. they aren't even from a location that was part of the Mycenaean world either. I do agree that it's very possible that parts of Greece, especially those outside of the world of the Mycenaeans, may have had much more Steppe admixture but it remains to be seen if this is actually true in the LBA/IA/Classical time periods. Iron Age/Classical era samples from Epirus, Thessaly and Macedonia are definitely needed to shed some light into the genetic structure of northern Greece.
As for your suggestion that this is the reason for the genetic shift in southern Greece, it is definitely a factor to consider, but I have serious doubts it is the main factor at play. More samples are needed to find the exact situation; the singleton from Roman Marathon to me seems more likely to be a Greek with roots from West Anatolia then a good proxy for the average Greek in Attica / Peloponnese. That being said, I do think most Greeks in the areas in question will be a mix of Marathon-type autosomal and the Mycenaean/Himeran-type autosomal, judging by the east-to-west population movements that were occuring in the Roman and Byzantine periods.