If the Keynesian approach can be mathematically derived from the Neoclassical one, that could only be because the normative aspect of both approaches is being left out as naturally unscientific. Yet that’s why the evidence is irrelevant here and why these respective economists will cling to those models for centuries regardless of whether their models have been falsified dozens of times. It’s because these aren’t strictly scientific models in the first place.

Underneath the scientific veneer is a philosophy about how society should be organized, and about which class of people should benefit most in life. Regardless of whether economics is a science, the field has been politicized like no other. There are moral aspects here that are left out of the economics textbooks because they detract from the scientific image that economists want to project.

Rather than respond at length here to your remarks on math and physics, I’ll write up an article on whether economics is a science, and maybe you can have a look at that. At any rate, I’ll try to take your objections into account.

I’ll just say here, though, that I think you’re mixing up simplifications and absurdities. Physicists simplify by abstracting away the interaction between systems, to focus on one system itself, which they often isolate in the lab by controlling the variables. Economists can’t do that, so they work with thought experiments (and with equivocal historical data and statistical fudging), and thought experiments are much more freewheeling since we can imagine any crazy scenario you like.

Also, there are lots of reasons to think that models in physics are scientific that don’t apply to economics. Thus, while that’s a valiant defense of economics, it’s not a terribly strong one.