I saw this map recently, and it got me thinking about this question, specifically with regard to the Balkans, but also to atleast Europe in general. At what point did fixed ethnicities and cultures really start to form in a way that their prospect for endurance was likely. For example, on this map, is it possible for all of the people of Dushan's Serbia to identify as Serbs and speak Serbian if the general extent of these frontiers could somehow be maintained? Would the identity of this state remain clearly Serbian if it ran as far south as Athens and the Morea?
The nationalities in Europe, at least in christianized Europe, are fixed for the western part between the VIII century and the X. We could see the appearance of differences betweens Franks, "Romans" (Aquitains or Italians), Hispani, etc.
As the germanic "nations" have more coherance and as the germanic languages formed themselves sooner than the roman ones, you have many determined ethnic groups in N-W Europe in the middle of IX, the romans in the late IX critically because of the turmoil in Mediterranean.
In Central Europe, the invasions of Suebi and Magyars, quite well coherant groups, help forming sort of "proto nation-states".
In the Balkans, it's far more complicated. The distinction between "Romans" and Greeks appears quite late, in the XIII and the shock of 1204 had something to do with that. After the appearance of separate states referencing to ethnies (the angevine Kingdom of Arania, by exemple), the situation is more clear, but the newcomers tend to identify themselves more according the territory (b.e., Macedonians, Thracians, Bulgars) than as a new people.
So, if it was expending toward West, the Duckas' Serbia could have made the Croats or the Dalmats referencing themselves as Serbs, it's more difficult to see it happening much bigger than OTL, in these borders.