Only 14 Levels - It's not exactly rocket science, but cutting off levels at 14 means much less room for power creep. It also means the meteoric rise some characters experience will be quelled due to no place to go after a while.
Limited High Level Spells - If you want to cast anything over 6th level (5th for clerics), you first have to pretty much max out and then research it yourself. Just the fact that those spells aren't on the charts is nice; it lowers expectations at higher levels of play.
Magic Item Research - Goes with the above. I think this is laid out very nicely, and is a serious cash sink for wizards.
Proficiencies - I'm not a big fan of skills in D&D, but these are binary abilities gained as characters level up. My favorite part would have to be the customizability of the base classes, especially if you use a little DM fiat. If a player wanted to play a paladin, for instance, starting him off as a fighter then allowing him to select proficiencies from the cleric list would pretty much do it. No other tweaks required. I like it a lot and might steal this.
Spell Signatures - Isn't this how it SHOULD work in D&D? Dungeon Crawl Classics goes to the extreme with this idea, ACKS not quite so much, but again, probably needs to be in a D&D-like game.
Assassins and Bards - Probably the best versions of these classes I've seen in a D&D-like game in a long time. I'm stealing the assassin.
Charts - Hey, I like charts a lot. While I'd probably never roll on these charts, their presence provides some good ideas.
Economics - Yeah, this game is all about economics or something, but I don't know...just sort of falls flat. It's not exactly poorly implemented, but for whatever reason I can't buy into the idea that there are 50 people/square mile, even if that was the population density of the "Roman Empire". The authors keep referencing Roman cities and whatever, but they must have glossed over the part of their history books where the Roman Empire was mostly confined to costal regions, large tracts of unconquered land left pretty much alone. Yeah, the cities were densely populated, but I highly doubt rural Gaul, for instance, had 50 fucking people per square mile. That's ludicrous. Plus, what does that really have to do with medieval populations, anyway? Rome went from literally millions of people to sparsely populated by the time the black plague rolled around; did they just forget that?
The worst part is probably the uniformity of prices. What incentive is there to be a merchant in this game if everything costs about the same in all parts of the game world? Is there a guild system not mentioned that is fixing prices or something? Is the teamsters union controlling flow of goods between cities? No idea, but it's unrealistic. I have NO PROBLEM with it being unrealistic; it is, after all, a fantasy game. But don't keep talking about historical stuff if you're going to ignore history when it doesn't suit you. There is an example given about buying a boat in a town. The DM can roll a bunch of times to determine availability and cost. It's not a bad system. My problem with it is just this: how hard is it for the DM to make that determination using their own brain? Again, not against charts and rolling, but that sort of makes the DM irrelevant if everything is decided with die-rolls. Plus, the dumbest part is that in the example given, there is no ship available for purchase. Okay, I can see that. However, if I roll into town, see a ship in the harbor and decide I want to buy it, I'll walk up to the captain and make an offer. He'll probably refuse. If I offer a million gold pieces for a ship worth 50k, do you think he'd still refuse? Fuck no he wouldn't. All that die-rolling removes the roleplaying element out of the game, which makes the economics system completely unlike anything in history. Well, it's exactly like modern times: I order shit from Amazon, and there is no haggling. Either they have the product, or they don't, and it costs a fixed amount. This is nothing like historical reality. Again, fine, it's a fantasy world.
I suppose this is neutral because even for all the gripes I have about the economics system, it's still Not Bad if you want to use it in a game that accepts the outlined premises.
Domain Rules - I have the same complaints about the domain rules as I do with the economics system, namely that they don't really model anything historical, but try to pass themselves off as such. Plus, lots of rolling to determine things; my high-level fighter cracked a lot of skulls even after he was king. He didn't sit around telling people what to do. I know the game has king in the title, but rolling a bunch of dice to handle problems is pretty boring. I dunno, the rules themselves are fine, but I can't see this as being very interesting to play.
Tax Revenue - "...is coin paid directly to the adventurer by his peasants." What in the literal fuck? I thought this was supposed to be rooted in historical accuracy or something. No one in fucking history sat on giant bags of money except Scrooge McDuck, and he's not even real. It doesn't say 2gp of "worth" or "value" or "services", it literally says coins. This is directly from D&D, which again is FINE if you want to say people actually used money instead of barter (they didn't), but hey, I thought you wanted something more historical..? This sort of thing makes it hard for me to tell the intent of this game.
Elven Spellsword - And Dwarven Vaultguard and Elven Nightblade. The classes themselves are just fine, but these names are gay as hell. When I was reading the class descriptions I was almost embarrassed about these. Elves are fruity enough, there was no need to go with the full-on faggotry.
Really Bad Things
Attack Rolls - This is the only deal-breaker in the game, but it's irritating enough to give its own section. There are two major ways to roll attacks in D&D-like games, and ACKS does neither of them. I'm not a huge fan of ascending AC, but rolling is super easy: d20+STR bonus+class attack bonus. I tell the DM what I got, he compares to the AC of the monster. Pretty easy. Descending AC usually means I have to refer to a chart, so I can either tell the DM my total roll and he looks it up, or I say something like "I hit AC 2". Either way, it works fine. ACKS, however, requires the player to roll over a certain number added to the AC of the monster. So the chart says something like +5. If the monster is AC 4, this means I need to roll 4+5=9 or over on a d20. What's the problem here? It's fucking stupid, that's the problem. In both of the previous two cases, it takes literally 1 second to determine if an attack hit; in ACKS if the DM decides not to tell players the AC of the monster, that's just more fucking math to do for absolutely no reason. None whatsoever. Everyone starts with an attack throw of 10+, why not just roll this into the monster ACs like 3rd edition does? Why the FUCK did you feel the need to lower AC and put this in the attack table? It's just stupid as hell. Arrg. Seriously, the only explanation can be that the designers wanted something "different", but really, it's just ascending armor class made more opaque. That is fucking worthless and dumb. I said I wasn't going to rant, but this is annoying.
Overall, with the exception of the idiotic attack rolls, ACKS is a decent D&D-like game that has a some good things going for it. As a game that tries to unify low-level play with high-level domain management, it leaves a lot to be desired. I browsed my old Mentzer sets when typing this up and cannot see any reason whatsoever to use ACKS over the system described in those books. What someone needs to do is either goad WotC into reprinting the Mentzer sets, or clone the domain rules outlined in those sets. $40 was a bit much to spend on ACKS, I think. The PDF is probably worth the $10, but print it out yourself and skip the hardcover.