I think the problem with simulating an economy (or a real ecology, for that matter) is that once the simulation starts, you're going to have to hope it stays within game-viable (playable) parameters. It's fully possible that a randomly generated economy crashes and burns within a few years, depending on what civs have what resources, the price of goods, and how wars work out . . . which makes things rough on fortresses depending on trade. A crashed economy might wipe out world trade almost altogether, taking civilizations with it.
This sounds familiar somehow. I guess there will have to be ways for them to restart the economy in the game so after a large crash, things start to climb up again. Until they crash again.
This would simulate the rise and fall, and restructuring of empires rather well. Plus, add in the fact that it would add some challenge (how WILL you manage your fort through tough economic times?) to the game which is what people always seem to be clamoring for.
We've shown that a single fortress can easily supply the basic needs (food, clothing, booze) for itself, so even with the addition of "tough economic times", the worst that I could see would be die-back and weakness to outside invasion. As long as settlements can supply their basic needs independently, you shouldn't have everyone dying off, which is fine. Droughts hit, larger cities starve, and the civilization rebuilds itself from the outposts. Or gets overrun by elves.
The problem that I can see developing is the typical dwarf fortress avalanche of crafts, where supply greatly exceeds demand everywhere. There would need to be some kind of use for things that weren't food/booze/clothing, and those things would have to wear out (otherwise, crafts become worthless). There might be something like "sites supporting 100 or more creatures require x crafts per year" or something like that that could work as an abstraction, but there might still be enormous oversupply issues. On the other hand, if civilizations could have resource shortfalls that could give them another reason to go to war with their neighbors, rather than pulling an infinite amount of star rubies from off-screen.
To risk thread derailment, the problem with free markets is that they are fundamentally greedy algorithms (favoring short-term gains for individual actors), which are efficient and easy but often non-optimal, especially over the whole group. In situations with limited resources and many actors, they also tend towards a single actor monopolizing all the resources. I don't think that a truly free market would work well in DF, simply because players wouldn't put up with being put in the position that a small late entrant in a 200-year old market would be put into. I think that the economy model would probably want to be similar to the late middle ages, with feudal regulation of a "free" market. Merchants set their own prices, but the King sets the taxes, can demand tribute, mandate that X be made for the common good, and so on.
For bonus points, make the economy flexible enough that it can head off towards other forms of government (constitutional monarchy, republic, anarchy, etc. to choose some historical examples) if the nobles are abusing their power too much. That's probably a bit too complicated, though, since it would require modeling a bunch more stuff.