IIRC Turchin also said that states on a major civilisational boundary ("meta-ethnic frontier", as he calls them) are more likely to form empires, because they have a clearer "them" against which to define their "us" and are therefore more cohesive. Since the steppes were unsuitable for most agriculture you'd still have a meta-ethnic frontier between the settled Chinese and the nomadic steppe peoples, although with less threatening northern neighbours there might be less incentive to unite compared to OTL.
Yeah, I think that's one of his ideas. IRC it wasn't important / mentioned in the imperiogenesis paper I remember reading from him though, which just used a simple model of agricultural expansion (centered on the old world centers of domestication and based around their [slightly questionable] models of how those spread) +"Miltech" expansion (centered on where those centers meet the steppe).
Intuitively "meta-ethnic frontier" seems like a plausible idea to me, but it also seems plausible that populations on frontier might have a more fluid and less definite idea of who they are as well (adopting customs from other people they encounter, switching subsistence strategy with climatic shifts and available land). Especially if it's not as much of a militarized frontier.
And the other difference with OTL might be change in the Chinese mentality. China might become less 'closed', more inclined towards expanding into the outer world.
I mean in OTL sending fleets somewhere far away and spending finances for similar purposes was kind of reckless. Everybody knows that this money and efforts would be much better used on maintaining the Northern border against the nomads - because just wait and in no time the hordes of mounted warriors would gather and storm China. And then your overseas lunatic ventures won't help you.
In the alternative "world without horses" there's less need to keep an eye on the North and that Chinese attitude might change...
Although if they did, I wonder what they would find. In a horseless / steppe nomad free world, the butterflies for the development of early civilization in West Asia / South Asia / Europe are enormous. Even if it's all the same until Rome, you would butterfly out the Hunnic population movements, later expansions by Arabs, etc. Alternatively, even earlier might avoid the Late Bronze Age collapse. The whole Western world that seemed millennia ahead until the Iron Age and still hundreds of years ahead (roughly) by the time of quite late Rome might have survived. Maybe the ATL's China would see merchant fleets from Indus Valley successors or Mediterranean states turning up early in whatever their equivalent time of the Han Dynasty or earlier (their own outgoing overseas ventures or no).