First, what I was proposing was not about preferences at all, but rather about how things are produced - if you want, you can just take the current consumption

First, what I was proposing was not about preferences at all, but rather about how things are produced - if you want, you can just take the current consumption preferences, or put in your own into the model. For example, the model could tell you what the price of drugs would be without all the marketing expenses, and I claim it would be (in the U.S., say) several times smaller, and as such we can measure efficiency of the system (or different systems) if we compare reality to this theoretical value. Here efficiency is just a measure of how much waste the real system needs (and this might differ from neoclassical economics definition of "efficiency"). Speaking of preferences, I think you're buying too much into Hayek's free market ideology here. I don't think was proven in any way, in fact as you state, if the definition of preference is that cannot be revealed in other way, then you're starting with that as an axiom. I would also say that I was born in a socialist system (with almost completely regulated market and planned economy), and it seems the last problem this system had was that the people's preferences weren't clear. In fact, if they weren't clear (for example, as queues for basic necessities) then you wouldn't be able to say that system was inefficient! Also, there was some internal competition, new products were created, and so on. The thing really was that the prices didn't reflect the preferences, which created lots of inefficiency (waste).