I saw this 'short cut' once.....

As in this example, the panels were next to each other, with but a short nipple connecting them. The nipple was secured properly, with locknuts on both sides of the sheet metal.

Unfortunately, in the thirty years or so since the place was built, the connection was no longer as good as it once was. A little corrosion, a little vibration ... and the resistance of the connection increased. In effect, the nipple became a resistor.

Since the neutral was bonded to the case, the nipple was the only return path for unbalanced current. With the nipple no longer reliable, the "voltage to ground" varied greatly as loads changed during the course of the day.

Finally, one Thanksgiving holiday, the 'swing' in voltage became great enough to fry a large number of small surge suppressors. Looking at the circuits involved, all were powered from the same phase - though several circuits were involved. We looked and looked and looked - then realised what we were NOT seeing was any neutral wire to that panel!

I repaired this, and all was well. Indeed, the IT guys were astounded in all the mysterious power quality issues that went away! They had more fun with all their fancy scopes and meters than a kid on Christmas [Linked Image]

Since then, I've be a lot more focused on keeping my ground path and my neutral separated.