toward a reflexive economics?

I hope (descriptive) economics as a discipline takes seriously its failures, which are probably a result of the limited scope of its analytical models, as well as the idealizations built into them. A kind of reflexivity that could resolve whether descriptive failures originate in one or the other of these causes would be a good thing, I think. But there are conflicting "desiderata": a loss of explanatory power (or scope of explanation) comes with any reduction in the type or degree of idealization built into a model. Part and parcel of work in the natural and social sciences is simplification. The ideal gas law abstracts from both molecular size and intermolecular attractions, and so is very accurate only for monoatomic gases at high temperatures and low pressures. 

Similarly, the model of "perfect competition" implies that in the long run firms can only make zero pro fit; but the model assumes, quite unrealistically, that a single firm's actions have no e ffect on those of others, unrestricted market entry, and zero product differentiation. I don't think economists have any delusions about the truth of these of assumptions. The difficult and subtle questions come with figuring out which ones to eliminate.