Serbs pre-Slavic is modeled with both Illyrians and Thracians and that should be realistic and in line with their position in Balkans.
Target: Serbian
Distance: 1.2317% / 0.01231690
32.0 (Balto-)Slavic:DEU_Krakauer_Berg_MA_KRA009
28.0 Balkan:BGR_IA_I5769
14.4 Balkan:HRV_IA_I3313
13.4 (Balto-)Slavic:DEU_Krakauer_Berg_MA_KRA011
8.4 Balkan:HUN_MA_Szolad_SZ43
3.4 (Balto-)Slavic:HUN_Avar_Szolad_Av2
0.4 (Balto-)Slavic:Moravia_POH13
with Montenegrins I removed Thracians, as they have no such blood. Slavic goes down to around 40% and fit is still exceptional.
Target: Montenegrin
Distance: 1.4128% / 0.01412840
35.0 Balkan:HUN_MA_Szolad_SZ43
24.4 (Balto-)Slavic:DEU_Krakauer_Berg_MA_KRA009
23.8 Balkan:HRV_IA_I3313
15.4 (Balto-)Slavic:DEU_Krakauer_Berg_MA_KRA001
1.4 (Balto-)Slavic:HUN_Avar_Szolad_Av2
aggregated
Target: Montenegrin
Distance: 1.4128% / 0.01412840
58.8 Balkan
41.2 (Balto-)Slavic
Target: Serbian
Distance: 1.5453% / 0.01545250
58.2 Balkan
41.8 (Balto-)Slavic
Serb G25 average is heavily Montenenigga shifted so Bosnian and Croatian, and even many Serbian Serbs will be more Slavic than that, up to half and above half.
But old models where some ppl get like 70% "Slavic" etc look to be completely off.
This figures I get are in line with models of that unreleased Reich paper. If there were Dalmatian Croats on G25, they'd probably model little bit above 50% Slav just like in the paper.
Croatian average on G25 from Zagreb isn't of much use in this because it's filled with heavily German admixed samples.

