I do not disagree with you. :smile:I think the code is totally wrong about the need for a GEC for a transformer that is located within the same building as the service disconnect.
Regardless of the 'elemental physics of the application' I believe the CMPs intent is to run a separate conductor for the GEC.Along with that, I think most if not all non-believers are so rapt up in trying to follow [or make] the letter of the "code", they forget about the elemental physics of the application and purpose.
As the code is presently written I see it as ambiguous, we will have to see what the 2011 brings. :smile: