Let me put this to rest. I measured the differential voltage between the ECU's ground and the point behind the glove box that the Vishnu instructions cites. With various loads (AC, light, etc.) I only found a maximum of 0.008V or less than one LSB.

This is not an issue, thus the grounding at the ECU bracket is ok.

 

See? Ground is ground.

Seriously though, thanks for the check. The Voltmeter I have in my garage is so crappy I wouldn't trust it to measure anything besides the existence or absence of a 12v potential.

d

 

Hey...I always try to admit when I'm wrong

An interesting find is that the ECU is grounded to a point just behind the battery where in fact the battery ground is also connected (the ground line continues on to another point that I didn't bother to locate.) That said when you measure differential voltages you (read your truly) should not use the battery ground post as there is a larger difference at that point. This is explained by the finite resistance of the connection between the battery and this first ground point, as well as a non-trivial current when the car is up and running...

Thus when I said there was 0.5v in the engine compartment grounds, in fact I was measuring the loss across the battery ground wire. So much for the grounding kit advocates....