The thread is about R1b-Z2103 (rather than R1a-Z93 or R-related lineages in general), and about its modern (rather than its ancient) distribution. However, regarding the points made, I have no opinion to express on whether R-related lineages had a North Eurasian origin. And there is no reason to suppose that various branches of R1a did not expand both inside and far outside of Iran over their very long periods of formation and development; to try to locate a single pinpoint is futile.
Data from modern populations provide a lot of detailed and accurate information that could relate to the most likely migrations of surviving branches. On the other hand, data from ancient samples is relatively sparse, commonly relates to lineages that have died out, and is frequently beset by classification errors. However, no information should be wholly disregarded; everything should be brought into consideration.
Regarding modern populations of Z2103 specifically, I do not see anything in the data to demonstrate a North Eurasian aspect to the development of its surviving branches. There is a cluster of related early branches that coalesces around Armenia and a similar cluster that coalesces East of the Dniester - these are the branches that survive and thrive today.
If anyone has any information to suggest that the immediate common ancestors of the Eastern branch of modern Z2103 most likely lived in an area other than the south eastern shores of the Black Sea around 3,000 BC, then I would be interested to hear about it.
Furthermore, given that (i) Yamnayan Z2103 samples exhibit a highly uniform and stable autosomal mix, and (ii) non-Yamnayan European Z2103 samples are very diverse autosomally (both from Yamnaya and from each other), is there any more likely explanation for this than that their DNA developed separately within different European populations at an earlier date, rather than as part of a single Yamnayan incusion at a later date?