ACKS has an economy which is self-consistent and scalable to different levels. You can skin that however you want, if your goal is to have an interesting end-game that mixes well with the early-game.
ACKS also puts a number of the initial assumptions under your complete control. For example:
Athas: multiply the cost of metal goods by x100, and use the centralized settlement pattern (2 rows) on p. 231. This—without any further work by you!—makes certain items harder to get in smaller towns, and produces large city-states while stripping urban population from the more difficult to sustain villages and towns.
Can you make it more complicated? Yes, but you don’t need to, if your goal is a plausible and flavorful setting with a fun end-game to match the early- and mid-games.
The post arguing against ACKS seems to imply that you ought to run a proper and correct economic simulation—for a setting for which there is no historical antecedent!—or give up. I think that’s ridiculous.
(I also think that An Echo solves the “some GM calls required” problem by saying, “Okay, now everything is a GM call, but we’ve made things simple and bland so the GM calls are easier to make.”)