It would be interesting to see if they could look at mitcochondrial DNA and y-chromosome DNA to try to differntiate between these two because I believe a common pattern in conquest scenarios is for the conquering men to marry the conquered women and eliminate the conquered men.

Stage 1) Cavalry army, 100% population A invade terriotory of 100% population B and say for the sake of argument the end result is 40/60 (not admixed yet, stratified).
Stage 2) Army now 40/60 (A cavalry and B peasantry) invade second population B further south with end result say 20/80 (stratified at the top).
Stage 3) Army now 20/80 (A cavalry and B peasantry) invade third population B further south with end result say 10/90 (stratified at the top).
etc

Or looking at it another way as an army of
- officers
- ncos
- grunts

With class-based assortative mating generally the admixture would either be at the same level i.e. grunt-grunt, nco-nco, officer-officer or (generally male-mediated) downward i.e. officer-nco, officer-grunt, nco-grunt so if population A becomes the whole officer class then there is no assortative admixture with population B at that level, only downward.

Say initially you have
- A senior officers
- A junior officers
- A ncos
- A grunts

After stage 1 you might have
- A senior officers
- A junior officers
- A NCOs
- AB grunts

After stage 2
- A senior officers
- A junior officers
- AB NCOs
- ABB grunts

After stage 3
- A senior officers
- AB junior officers
- ABB NCOs
- ABBB grunts

(Basically i think admixture models might vary between conquest and infiltration scenarios – for want of a better word to describe farmers hopping from defendable site to defendable site among a sea of HGs – with the first event in your model being an infiltration scenario and the second a step-wise conquest scenario.)

brahmins sound of the vindhya are pretty much the same. there’s no cline that i see. so it seems like the ashkenazi jews: early admixture => later radiation.

The spread of a cavalry culture would have to come after cavalry-sized and numbered horses became available at some later time being particularly significant in places where everyone using horses would increase their net food-getting capacity as opposed to those places where keeping horses specifically for cavalry literally ate into their net food-getting.

The places where everyone used horses would then have a significant latent military advantage any time the people in those places united in one gang – at least within the logistical range where large cavalry armies and all their spare horses could be supported at the time (protecting western europe and southern india a bit maybe?)

In terms of your model i’d see the first farmer types spreading out from Anatolia down through Iran and/or fertile crescent (i personally wonder if they were blocked by the locals at the Indus valley and mostly spread around them via the coast) or Indus valley plus the coast. Either way the south would become coastal farmer / fisher settlements with a native HG interior that mix over time.

(Time is the critical element imo. If the first farmers could have spread all over without any blockage i think they would have and almost everyone would be descended from them.)

Then later you get the scary horse guys arriving in North India and smashing things up giving the second mix. Later again kingdoms containing a significant proportion of scary horse guys push south in a step-wise fashion with the proportion diminishing with each step south giving the third mix.