This quote has always seemed cute, and coming from Berra it would be, but hardly true.
Theory (at least in physical disciplines) is always taught with up-front statements of assumptions, what non-idealities are assumed away in order to tractably develop a theory in the first place. It's not taught by theory that theory transcends whatever simplifications were made to arrive at the the theory. Physical theoreticians do not contend that theory has no differences from reality.
It's not a boolean matter of the assumptions being "not valid" - sometimes, the non-idealities are of significant enough magnitude to make a meaningful difference. Other times not.
If one is taught ideal spring-damper theory, they are told friction in the damper is neglected in the mathematical model. If I use this theory to size the spring and hydraulic damper for some application where the forces involved vastly outweigh the damper seal friction, it's likely this non-ideality doesn't impact the answer enough to affect my sizing.
If I'm trying to eke out every last bit of performance from the spring-damper system and minimize damper lag, then the seal friction probably does matter.
Either way, when the theory was taught, it was done by stating what assumptions were used to derive the theory. How much those assumptions affect the validity of the model for a given purpose is an "it depends" matter.
Theory (at least in physical disciplines) is always taught with up-front statements of assumptions, what non-idealities are assumed away in order to tractably develop a theory in the first place. It's not taught by theory that theory transcends whatever simplifications were made to arrive at the the theory. Physical theoreticians do not contend that theory has no differences from reality.