The assumption behind enterprise modeling is that "enterprise" is informationally sound.

This is a false assumption.

Much of what goes on within "enterprise" is done to compensate for informational failures often as a legacy of pre-computer pencil and paper limitations. 

Automating how we compensate for a problem in effect automates the problem.  If we compensate for a broken leg with cast and crutches, continuing to use cast and crutches after the leg has healed keeps us just as restricted as we were.  This is the essence of enterprise modeling; to "model" legacy limitations that could be "healed" by using both information and computers to solve problems instead of to compensate for them.  

By failing to understand information, enterprise modeling is just another way to expedite information problems instead of finding real solution.

The question that is almost never asked when looking at new IT techniques, methodologies, and perspectives is,

"How does this concept help us manage information better?"

The answer invariably is,

"It doesn't."

Enterprise Modeling is just another IT concept that avoids the real issue of informational failure, thereby perpetuating the problem.

To really help "enterprise" we need to improve information management so "enterprise" can shed itself of anti-informational practices that consume resources and bog it down.