I haven't been able to settle on an implementation of the cleric domain game yet. Running the numbers on how much divine power a realm generates is pretty straightforward. Garrisons of paladins are easy, and providing an NPC cleric to cast spells for the party is trivial by demographics of heroism. The real trouble is "what do you spend divine power on?"
Magic item creation is clearly in-scope, but if the highest-level NPC cleric in the realm is below domain level, it's going to be unreliable and kind of lame. At 500 families under church administration, you get a 5th-level cleric, who is 12+ on magic research throws, and 500 divine power per week (did I mention that divine power is hideously productive?). He can attempt to make a Potion of Cure Light every week; 2000gp/mo input materials, in expectation 0.8 potions of cure light output. Ehhhh (and that's assuming he'll sell to you at price to produce; NPCs in the domain are friendly, but not henchmen, unless so hired, which is much easier to do before they're in charge of things). Further, figuring out aggregate potion production for deeper trees in larger domains (where you have more than one cleric of 5th+ level, or assisting 1st-level clerics) looks to be a pain in the ass. Plus, high-quality item production really feels more like a dwarf / elf thing, or a wizard thing, than a church thing.
Ritual magic is really more like what I want, particularly Harvest, but smaller. Bonuses to feudal troop morale ("Deus Vult"), burgher tax income ("Give Unto Caesar"), population growth ("Go Forth and Multiply"), extend supply caches during siege ("Seven Lamps"), reduce fortifications (Jericho), rain locusts and frogs, call or prevent plagues, part seas to march armies across, drop prophecies, summon outsiders to mass combat, ... but at this point we're looking at a whole new subsystem of magic that has to balance size of effect and divine power cost and some sort of casting mechanism (I dislike the magic research roll mechanic, because it's very unreliable at low levels, but I do like the idea of rolling for divine favor, because deities are capricious). I'm tempted to just make these spells that require normal cleric spell slots, but also cost divine power over time, with a Divine Favor 2d6+wis reaction-style roll with results including "works great", "you're cursed", and "demands more goats".
So you see why I might be bogged down.
So at a wag, we have roughly three "tiers" of play; 1-4, 5-8, and 9-14. It might be worth breaking 9-14 down further, because scale increases quite a bit over those levels (company vs legion scale, for example). Then we cut our rituals into those three tiers - least, lesser, and greater. Least rituals are available at 1st level, lesser at 5th along with consumable magic items, and greater at 9th (er, 11th? I dunno, my games never reach that level anyway). Some rituals we can reuse across levels and change their scope - Least Deus Vult at 1st level affects up to a platoon (or so), Lesser Deus Vult at 5th up to a company, and at Greater Deus Vult at 11th up to a legion, with divine power costs growing accordingly. Your realm-economic miracles like Give Unto Caesar, Go Forth and Multiply, Good Harvest, or whatever have the number of families that they effect capped based on caster level, and their divine power costs scale similarly.
This is a fine idea, I guess, but it's mighty fiddly. Hard to strike a balance between our goals of "clerics can do something fun and interesting and relevant to the domain as a whole", "keep it simpler* than normal ACKS", and "make domains at lower levels viable."
The straightforward solution is to make realm-blessing binary / boolean; either your whole kingdom has Deus Vult up, or it doesn't, and the cost in divine power is per-season based on total number of Feudal families. Then your one-off miracles like calling outsiders have flat / fixed divine power costs per casting (and maybe minimum caster levels).
A church family generates 1 point of divine power per week (12 per season), while a non-church family generates 1 point of divine power per month (just based on their tithe expenditures; 3gp per season), so in a balanced realm of the four main estates we'd expect 21 divine power per season per four families. We don't want it to be possible to have all the estates blessed, because that takes the choice out of things (also: bookkeeping), so presumably we want costs of 7? divine power per family per season to bless an estate with eg bonus morale (I like the idea of this applying to estate loyalty rolls as well as troop morale, so it becomes useful for politics as well as war), bonus taxes, population growth, or protection from disasters (plague, city-fire, hurricane, drought, wizard-monster escapes dungeon, ...).
(A more ACKS-consistent solution would probably make heavy use of forum esoterica, like this and this)
Amusingly, I actually expect wizards to be easier. The ACKS domain rules work mostly-fine for wizards; they support doing wizardly stuff very well. You dump your money into libraries and workshops and every season there's a chance that a randomly-rolled new spell is added to the archives, where your PC mage can then learn it, and the domain provides wizard-troops during war. Very straightforward. Nobody expects a 5th-level wizard to turn a domain-scale battle, but a 5th-level cleric, is they prayed hard enough, conceivably (non-disbelief-suspendingly) could, because deities are powerful. And that's what makes this tricky.