Ahhhh. jseah is doing a wonderful job of selling the idea, as well. I've been such a motive force behind this topic that sometimes I feel more like I'm a salesman making a marketing pitch than a contributor in a collaborative idea.
Let me take this back up a level of abstraction for a moment, however, and talk about complexity and player control.
One of the real problems of the "complex but easy" nature of this game is that players generally have far too much direct, precise control over their problems. Need a steel battleaxe? Order one smithed. Need steel? Order it made. Need fuel or iron ore? Go designate something to be mined or deforested.
All of these "problems" are not real problems because their solutions are so utterly direct and simple. The player has the ability to solve the problems they face with pinpoint precision, and no fear of any sort of unintended consequences for sloppy behavior.
Consider instead trying to get mechanical automation of some sort of function in a fortress, however.
Repeaters are elegant, indirect solutions to a problem of trying to automate a simple action of pressing a switch on and off over and over again. What makes these systems neat and interesting is that you do
not directly control them. Even more advanced, see
Computing - an orgy of using simple tools in obscenely complex ways to solve fairly simple problems, but in extremely indirect ways.
You see, giving players indirect means of solving their problems (or rather, taking away the easy, direct means) makes the problems much more compelling and interesting even without making there be some sort of conflict of choice in the solution to the problem itself. The enjoyment of the problem comes from the simple ability of the player to react in a plethora of new and creative ways to try to find a better, more efficient way to solve the same fairly simple problem, because the problem cannot any longer be as simple to solve as just pushing the "solve this problem" button.
If we have a multitude of factors which all conflict with one another in an unstable equillibrium, and trying to manipulate any one factor of this equillibrium will inherently have unintended consequences with a half dozen
other factors, then the entire act of solving a fairly simple problem can potentially become a much more dramatic and compelling problem. Forcing someone to choose between two conflicting desires is the start of a decent choice. Forcing a player to chose how they balance a
dozen competing desires, and how they can prioritize the spending of limited resources to satisfy multiple different conflicting needs creates a real source of difficulty to the player.
This is partially why I'm trying to throw not just one system, but a
giant jumbled mess of systems that involve all aspects of industry related to food, fuel, textiles, medicine, trade, domestic happiness, and the sheer labor management of the fortress so that everything competes for the same limited resources you can divide between them.
And then I've got my other suggestions, such as
Class Warfare which proposes how to increase the complexity of social dynamics in the fortress, so that dwarves demand access to luxury goods, finer meals and foodstuffs, entertainment, and social justice. This means that you'll need to address these problems indirectly using the same materials and taking away some of the resources used for the initial problem of farming.
This is why I consider it the groundwork of all future expansion of what Dwarf Fortress can be - it has to start moving away from simple, binary solutions of "push this button to solve this problem" before it can start to really add depth to its complexity.