I recall that the cast-commentary on Serenity described the ship as "another character", and we see this reflected sometimes in Traveller. There was also a saying, back in Tribes 1 (I swear I'm not actually this old), that the missile turret on Raindance was the 11th man on a 10-man team.
Got to thinking, maybe we ought to treat domains more like we treat ships in Traveller.
In Traveller, the party as a whole has a ship. Sure, there is one guy who, practically-speaking, is in charge of managing most of the paperwork. You can get by with just a pilot, but in practice you want a pilot, an engineer, an astrogator, and a gunner or two for the whole thing to work.
What if domains were like that?
The party as a whole has a domain. The player with the highest paperwork tolerance is probably in charge of dealing with most of the numbers, and you can get away with just one character running it, but in practice you really want a full party, fighter-wizard-thief-cleric team to run a single domain.
This approach is somewhat internally-consistent with ACKS' assumptions; if you assume that most domains come into being via adventurers clearing territory, then you'd expect to see that enshrined in their origin myths and the structure of their institutions. The normal way for things to be run in this world is a council of oligarchs, one from each of the Four Estates (nobility, burghers, church, and wizard's tower - yes, I am absolutely stealing from EU4's Cossacks DLC. It's a good idea.). Moreover, each estate expects its representative to be independent, not a henchman of some other estate's representative. Per The Art of Not Being Governed, a fifth "hill tribes" estate willing to follow explorers, barbarians, and shamans might be worth adding. A domain's dwarven population might want a dwarven oligarch; likewise elves.
Estates have morale/loyalty scores instead of the domain as a whole having a morale score. Loyal estates give you good stuff, mostly usable by the classtype that is supposed to lead them (eg, divine power, arcane research stuff, troops, offensive hijink capability / reach), while disloyal estates cause trouble (the church foments peasant uprising against reprobate leaders, the burghers hire assassins, &c). You also need multiple estates cooperating to keep things running smoothly - nobles for garrison, burghers for cash+market, and divine for disaster-resistance, at a minimum. As a result, having a party split up into many tiny personal domains means that each will have a great deal of trouble, while uniting as a single domain reaps rewards.
Might also be worth assigning estates a strength score, affecting their ability to cause trouble as well as their ability to provide benefits; the priests cannot foment a rebellion if that have all been killed, but likewise there will be nobody to buy a Restore Life and Limb from (market class for divine spellcasting reduced to VI). Strong, loyal burghers pay lots of taxes, but strong, disloyal burghers have lots of money to hire assassins (per-season per-estate random event rolls, modified by estate loyalty, anyone?).
The ACKSiest approach would be to track estate strength numerically with families. I recall reading that historically, about 90% of the population farmed, supporting the remaining 10% who specialized. This is reflected in the default assumptions about the urban population fraction in ACKS (though there it's actually 90.9% and 9.09%, which is a pain in the ass). Taking the "no land area, only families" simplified domain approach I postulated here, when you gain families, you place them under the administration of one of the estates, which is then responsible for extracting labor, taxes, and such from them and yielding some benefits to its oligarch and the realm as a whole.
The noble / feudal estate provides a knight and his retinue for every 10 families under its administration (the ACKS-math works, with demographics of heroism providing 1 2nd-level character per 50 people, which is 10 families), as well as providing garrison for the realm and labor and materials to build and maintain strongholds. The burgher population determines market class, as well as providing labor and maintenance for a navy, which generates trade income in peacetime and can transport and supply military units during wartime. Families laboring for the church generate divine power, which can be used for realm-scale blessings, averting disasters like plague, and personal research. They also support paladins, although at a lower rate than feudal families support knights. Wizards are trickier; we want to both provide a benefit to the domain as a whole, as well as to the mage-oligarch personally. Monsters and magic items are hideously expensive and not super effective, but if ever there were a time and place to train wizards in mass-combat quantities, it would be a prince-wizardric (existing within a larger domain, like the prince-bishropics of the Holy Roman Empire). Such an organized enclave would also maintain libraries and labs, and possibly a repository of spells. The hillfolk are unruly and pay very little tribute to the realm, instead redirecting their economic surplus into population growth. They're a good source of light/skirmish mercenary units, though. There are a lot of things dwarves could provide, including siege weaponry, heavy infantry, fortress maintenance, high-quality nonmagical equipment (gunpowder?), and loans from the vaults. Elves have mass-combat spellswords and top-tier archers (possibly spellsword archers?), and are a source for monstrous mounts (giant eagles, gryphons).
I guess one of the goals of this, in addition to bringing the party together on one domain by design and eliminating the need for hex-mapping, is to make domains give you tools, weapons, stuff-you-can-use, rather than just cash. All too often in previous domain play, players had cash but needed mercs. Having families provide less cash, but also obligated troops, helps address this problem. Also, cash is boring, troops are fun. Eliminating hexmapping also opens up more possibilities for the low-level domain game, because you no longer have such strict requirements for clearing hexes and claiming land. If you're third level and you have ten families in your 30-acre "domain", you can start getting troops out of them very early. Further, switching to just tracking families under nebulous governing institutions removes the need for deep NPC-trees. It's absolutely unrealistic for modeling Medieval Europe, where governance was intensely personal, but man it would make life easier for me (provided a decent spreadsheet to manage all the stuff that domains provide - then it's just "they're a very militaristic realm of 10k families with medium priests, weak burghers, and no wizards, call it a 50% feudal, 30% church, 20% city split on population, so that gives 500 knights, 500 squires, and a thousand pikemen led by an 8th-level fighter, 150 paladins led by a 7th-level cleric, and a class IV small city market run by a 7th-level thief, with a fleet of one small galley and three large sailing ships. Boom done, go invade it already, and take their families for yourself."). It is also a better fit psychologically for my players (and possibly non-historian players more generally), who typically expect impersonal states with a couple of figureheads at the top and aren't huge on the courtier game.