IUltimately neither the Vlachs nor the Albanians ended up being a Slavic group, so I don't see why the period statelessness between the 7th and 11th century was necessary or more likely for non slavized states to appear.
I think I see our disagreement there : you look at modern communities of Albanian and Romanians and postulating that, well, they survived so they weren't really in danger to not to. Thing is we know that,
@althisfan posted it, that while the modern Albanian population alltogether is rather close genetically to ancient inhabitants, that a good part of Epirus was acculturated. The number of Valach communities that neither made it to being states and were swallowed up and acculturated between Middle-Ages and Modern history is staggering. As it happened with Britain, you have a whole Romania Submersa there.*
At some point, especially for what could be considered chiefdoms and early states, a too great expansion over different (culturally wise) groups is risking certain acculturation even in a dominant position : it's what happened to Bulgars, and probably what would have happened to Avars.
More than that I don't get why the distinctiveness of the Albanians would just appear like that without any particular reason.
As mentionned above, linguistical distinctiveness played a great role : there's a thing having difficulty to separate two close southern slavic communities on this basis alone, especially with a same material culture, than proto-albanian and slavic even sharing the same material culture. Of course, the way-of-life is as well particular enough : highland communties tend to distinguish themselves from lowlads trough conservative and relatively cloisonment as it happened, for exemple, with Kurds, hughland Scotland, to say nothing of the mess Caucasus is.
At least considering just the Highlands, I'd say that a state coming out of this region would unlikely to become Slavic long-term.
Any Albanian state will have to devellop itself outside the highlands to be viable, would it be to gather enough resources to be mobilized in a state creation phenomenon. IOTL, it was quite gradual. Again, there's no thousands of explanation why Valachs managed to form viable chiefdoms and early states essentially in the Danubian basin, and not in the Balkanic highlands. An Albanian state that would be limited to its immediate highlands would be likely to end up lie Byzantine Vallachias, as in autonomous enough but extremely dependent from the primary state for their survival.
I agree that my approach thus far is quite structuralist, but I think that's a model that can really help understanding the state-building process.
But why did Albania apparently become less mixed by the late 11th century?
It didn't? Again, genetical and cultural mix are two different things. Most people using English in Africa aren't much of Anglo-Saxons either.
More to the point, it happens that Albanians lived in a peripherical area of Romania and Bulgaria alike, and seem to have enjoyed an autonomy large enough during the Xth to XIIth century (maybe not fully Christianized until the Xth), and eventually this autonomy may have been "fossilized" in the autonom Principalty of Arbanon (which possibly covered an earlier political ensemble led by Byzantine Albanians, coming from the struggle against Italo-Normans, IMO).
If anything, Albanian probably got more mixed after the late XITh with the growth of Albanian statelets and migrations to southern Epirus and Greece. The key point, there, is that identity comes, in pre-Modern times, more often from political/institutional context than the reverse.
I guess what I mean is that things seemed to be directing towards a creation of a separated Albanian group(maybe by virtue of a local dominance of Albanian groups) rather than this mish-mash of Albanian, Slavic and possibly Aromanian groups just ending up with IOTL Albanians,
I tend to think that the discussion we had with
@althisfan points that the distinction is a bit moot. Every community is mixed up to a point, and he was right to point that Albanian seems to have been less mixed up, the difference possibly being between "highland" proto-Albanians and "lowland" post-southern Illyrians IMO, altough the difference is quite artificial, as it's likely that the seconds might have shared linguistical closeness up to a point.