Careful mentioning any actual economic theory here lol.

I remember being annoyed on the 2nd day of Econ class when the prof mentioned that empirically we know human preference graphs can be cyclic but we have to pretend that's not true for any micro to work at all.

That's succinctly phrased. I don't know why it would bother you. It seems like it's admitting a limitation of the model he's about to teach.

It's a fundamental premise of the entire discipline and it's known to be empirically false

Does it also bother you that engineers designing bridges use Newtonian physics and discount relativistic effects?

There's not a scale at which acyclic preferences is a good approximation of what actually happens, so that's not really a good comparison. But even given that, structural engineers use nonlinear models that are more complicated than Newtonian physics, whereas economists say "math is too difficult so we'll use a simplifying but false set of assumptions"

Apparently I choose a great example then!

There are a lot of examples where the assumptions, especially this one, of microeconomics hold. Profit-maximizing corporations don't have cyclic desires. In many cases, people don't have cyclic desires (e.g. in the flavor of the soda they want). Microeconomics starts with those, much as becoming a structural engineer requires understanding the basics of Newtonian physics.

They get into more complicated math in more advanced courses, but they don't start there. All of engineering, science, etc, is "the math at a deeper level is too complex, here's a simplifying but false set of assumptions". And then a subset of people who learn about that care about cases when the assumptions break down and studies the next level.

Right, but Newtonian physics can fairly accurately predict gross body mechanics whereas economics can't even predict the large-scale wage or price level. They'll simultaneously model the income effect and the replacement effect of taxes and then shrug and say "of course we can't say in advance which one will dominate".

"Wrong" isn't a boolean. There is a vast spectrum of wrong, and Newtonian physics is so unwrong that you can go to the moon with it. Economics is so wrong it has never worked.