One point which I haven't seen made here is that perhaps economics is a reductionist approach to something which cannot be easily modelled in a reductionist way

One point which I haven't seen made here is that perhaps economics is a reductionist approach to something which cannot be easily modelled in a reductionist way. Nobody here expects everything we know about Physics to become largely irrelevant tomorrow. But deep within is the assumption that the underlying models we build today will hold true tomorrow. An electron will always be an electron. Now let's look at biology: a frog today may not be a frog two-hundred years ago. Microbiologists are finding new species that simply didn't exist 50 years ago. But there's a shred of something fundamental there at the physical level: a reaction involving biological building blocks can be deemed to be as true as hundred years from now; whether or not organisms develop more stable sets of chemical reactions. Psychology and social sciences are changing because people are changing and their environment is changing. Pretty much the only thing that remains constant is that people have fundamental needs. As people become aware of cognitive biases, and more are exploited by them, suppressing them becomes not just an issue of utility but of survival. Yet somehow, we expect economics, which in many ways is the study of a very dynamic system which builds on social sciences where the underlying rules are changing and the very reductionist terms we are using are changing, to yield results which are accurate and work at all times. Can we build better macroeconomic models? Almost certainly. Will they be valid in 100 years time? That depends. If we think to Physics, Thermodynamics gives a field that seems to be universally true -- if you come up with a model that violates the second law of Thermodynamics, it's fair to say your model is wrong. It doesn't matter if you're studying distant stars or subatomic particles. At the same time its predictions are somewhat vague -- we can derive upper and lower bounds on things, and say with an almost certain degree of belief what cannot happen and use this to derive models that are consistent. We don't seem to have much like that for economics, although if we focus on resources we can clearly see there is a physical upper bound. That seems like it might be a better direction to approach the problem in than from the end were even the idea that employers employ people to create more value than the money they pay them cannot be taken to be universally true. Of course, nobody has any incentive for denying most Physics, with the exception of the 2nd law of thermodynamics -- because there is immense money to be made by charlatans promising infinite energy in exchange for lots and lots of money. Given economics deals with matters that can have direct implications for those with lots of money, we would be naive to not assume the same to have happened there.