Ancestry Models

In Lazarides et. al. 2016, https://www.biorxiv.org/content/biorxiv/early/2016/06/16/059311.full.pdf,  modeling ANI (Ancestral North Indian)

While the Early/Middle Bronze Age ‘Yamnaya’-related group (Steppe_EMBA) is a good genetic match (together with Neolithic Iran) for ANI, the later Middle/Late Bronze Age steppe population (Steppe_MLBA) is not.

However, in Narasimhan et. al. 2018, Steppe_MLBA is seen to be a better match, and Steppe_EMBA is ruled out.  Underlying it are different other components of ancestry that are used.

In Lazarides,  the other components of ancestry are Iran_N(eolithic), Onge, and Han.
In Narasimhan, the other components of ancestry are Iran agriculturist (same as Iran_N?) , AASI (another name for Onge), and a composite of "Indus_Periphery" and Swat SGPT and early historic individuals.  

Why is it important? Because if these genetic findings are correlated with language (a big if!) and  because if the Rg Veda is taken seriously, e.g., its mentions of the Saraswati River, then a post-IVC arrival of I.E. is untenable.  Likewise with the Vedic Indra-Varuna-Mitra- Nasatyas in the Mitanni documents.  Steppe_EMBA doesn't rule out a pre-IVC-collapse arrival; but Steppe_MLBA pretty much does.

One should note that Steppe_EMBA itself lies on the cline between Iran_Chalcolithic and European Hunter Gatherer, and Steppe_MLBA is largely Steppe_EMBA but on the Steppe_EMBA - European_MidNeolithic_Chalcolithic cline.  (see Lazarides 2016 for this, excerpted diagrams here).
But in contemplating the difference between Lazarides and Narasimhan, it seems the Siberian Hunter Gather component is what makes the difference. 

A question would be - do Steppe_EMBA and Steppe_MLBA have "similar ancestry profiles"? Apparently in the context of the Narasimhan et. al. model, for the purpose of modeling ANI they do not.  Yet, for the purpose of modeling ANI,  Indus_Periphery and Swat individuals have a similar ancestry profile.