You seem to be arguing more that techs can require certain polity-specific conditions. I entirely agree with that - in fact, that was the second point made by the previous post you quoted. What I disagree with is that size by itslef is a particularly large advantage.Economies of scale. A big chunk of our techs are outright useless to a smaller polity, and only really usable at all if you are at Thunder Horse or Khemetri sizes. Another chunk requires resources specific to certain biomes, or the skill and knowledge infrastructure to copy it.
Take the Sacred Warding here, anyone can see it and try to copy it, but only the Khemetri have a decent chance of pulling it off without fucking it up. It takes a large priesthood devoted to scholastic pursuits to imitate.
Or take the volley archery, anyone could copy it, but they don't have the social foundation of large numbers of militia archers with an obssession towards perfectionism trying to perfect their shot, or the cultural attitudes towards unity to synchronize effectively.
The examples you cited are Sacred Warding and Volley Archery. Sacred Warding requires a well developed priesthood like you said, but it doesn't require size - in fact, I think The People were magintudes smaller than they are now when it was first developed. Volley Archery requires specific social conditions (large militia, perfectionism, unity), which again doesn't require size.
You've actually yet to cite anything that actually had requirements of scale.
I think you are vastly overstating the power of PiA here. We don't instantly get the techs around us, and they certainly don't instantly diffuse to our neighbors. At best, the channel is competitive with direct transmission; there is no way it is going to be dominant.And yet, when we got a direct route to the Khemetri, the trade still flowed through the Trelli, because they are the faster channel.
The same principle applies here with techs. Whoever in the region gets a shiny, we get it, then people copy it off us.
But we have the traits that make reverse engineering foreign ideas fast and effective, while they don't.