I love being an Enterprise Architect. But I feel there are some issues with the common approaches in EA. 

 For the most part, I think laypeople see this EA stuff as either an expensive science fair projector black magic that can only be practiced by the high priests.

This is a problem. 

The first aspect of this discussion will be explaining the value of a good EA program and what should it look like. And lastly the pitfalls in many of the popular EA approaches. Now I am not the highest authority in the land, not by a long shot. And I know that there are many variations of this conversation, but I believe that if the EA folks only ever produce stereo instructions and glossy crayon drawings then failure is in the cards. It almost seems as though it never gets beyond an academic exercise.

If you think about EA more as structural architecture then I think we can get closer to a better value proposition. The product(s) of EA allow the business side the ability to interact with complex technical environments and help facilitate good decision making by management. But it should also allow your techies the focused details they need with an explanation of the network tropes.