The danger is that you end up with a "perfectly spherical cow" that you base your arguments then on, and your abstraction-based results then have no bearing whatsover to the real world.
I'm not saying that you must follow your abstractions blindly. At the end you must always test your theory against reality and adjust if reality contradicts your assumptions.
The danger is that you end up with a "perfectly spherical cow" that you base your arguments then on, and your abstraction-based results then have no bearing whatsover to the real world.