I don't like situations when a two different branches of sciences discussing the same subject give contradictory results. This mean that there is an issue here.
From archeologic point of view the transition from Lchashen culture to Yervanduni era in modern Armenia is well defined.
First there is a major discontinuity with the apparition of Urartian culture after the 785bc. New culture is evident in sites like Erebuni, Argishtihinili, Teishebaini (Karmir Blur ).
While local Lchashen people who do not live in Urartian forts continue to live with their culture. Obviously they are influenced by the politically dominant culture. Which indeed is visible in their pottery, and not only the pottery. ( See the link below ). Later when the Urartu rule disappear, nothing remarkable happens. As far as I know there are no particular changes in material culture when shifting from Urartu to Yervanduni. Offcourse there are those destruction layers that some date at 640bc. But the destruction could be caused by internal turmoil or some other events that do not require new immigration.
This is what has to say the archaeology. Now let's see what is saying paleogenetics.
We have two aDNA from Teishebaini. As expected one of the samples has a typical Vannic Urartian profile. I13035. Very low Steppe and the highest Levantine shift.
Then we have numerous aDNA from Bover, Bardzyal, Harjis and Pijut. There was no data about Pijut but the rest are qualified as LM (Lchashen). Their DNA is fully consistent with their archaeological classification. Their DNA is virtually the same as in pre Urartian period which means that their settlements didn't witness any new migrants. The case of Harjis is even more ironic. This site is in Syunik which was not really controled by Urartu. So this would be the last place to find a Vannic DNA. Moreover they are marked as post-Urartu.
I think it's quite clear why we don't see any important shift in Urartian period. That's because there are no samples from densely populated Urartian sites in Araratian plain and elsewhere. What we have are Lchashen 6 sites in Urartian period.
After the fall of Urartian state the mobility of people living in those forts might have increased and gradually we will see them elsewhere. So I seriously doubt that there was any abrupt genetic shift. Rather gradual which is visible in Artaxiad era samples still having the archaic Etiuni genetics.
I hope another paper will be dedicated to this subject, where we will see more samples from true Urartian sites in modern Armenia.
PS There is another outlier with zero Steppe from Keti. I17184. This sample has the highest CHG coupled with high Anatolia. This pattern is typical to modern Kartvelians. Consistent with it's unusual autosomes it has G2a1 y dna which is popular in modern Georgia and Ossetia. So in most likelihood we see a person of Kartvelian? ancestry in Keti during the Urartian period.