I'm hardly an uninvolved onlooker in this, but I've had some time to sit down with ACKS and give its domain and campaign region creation sections a proper read. I think it's a very interesting book that does a good job of accomplishing what it set out to do- but it's interesting to me that it and AER have almost diametrically opposed ways of getting to their destinations.
The key structural difference between ACKS and AER, in my reading, is that ACKS attempts to directly anchor domains in plausible extrapolations of regional population densities, markets, and gold piece values. The roots of the domain system are embedded in tallies of peasant families per hex and a gp-per-family domain income value. Population growth, feasible land carrying capacities, and racial modifications for elves and dwarves are factored in to produce plausible population totals, and those are used to derive the lord's per-turn income. Out of this income, the lord pays stronghold upkeep, garrison costs, and other fees- including the presumption that he's paying 10% in tithes and a 5 gp per family festival cost at least four times a year.
Stronghold construction and settlement building are also established on a strict gold-piece basis; if you want a stronghold, you need cash, and a substantial amount of it. Prices are given for everything from a 10' stretch of corridor to moats (filled or unfilled) to the cost of a stone door. Everything is calibrated to a particular scale of gold piece costs. The plain intention is that as long as your campaign shares the default ACKS economic scale, the prices provided for domain operation should make sense and be useful in play.
The model is explicitly feudal. Your PC can't personally control more than one domain, while the mightiest empires on the globe might be composed of tens of thousands of domains. The emperors of these realms subinfeudinate their best domains to their henchmen, who in turn subinfeudinate all the way down. It patterns very strongly on traditional Western European historical models, right down to the assumption of tithing, festival costs, and duties/favors granted to vassals.
In summary, ACKS handles domains by extrapolating upward from base population numbers and handles their activities and influences through direct gold piece expenditures, income, and population adjustments. AER approaches the matter from the exact opposite direction. I think it has to do with some fundamentally different things that the creators of ACKS want from their games as compared to what I want from mine. ACKS does a great job of doing some things that I, personally, wouldn't ever want to do with my game. Which is all for the better, as it's in everyone's interest for new material to cover as wide a range of playstyles as is possible.