There is an inspector in our area who is writing a violation to the contractors using the large capacity non-metallic boxes for romex jobs where the box is ±3 inches deep. We are talking about our 22+ cu. in. boxes. When these boxes are nailed to a 2 X 4 stud they take up almost the entire stud width. The violation according to him is he says the box needs protection from drywallers so he wants nail plates installed behind the box. His code reference is NEC 110.27(B). I disagree because this code reference has to do with "Guarding of live parts" IMHO, he is going waaaay overboard. What say you?

I almost agree on the danger. I understand the concern if they are putting rock on the back side of that wall, I am just not sure the code actually backs him up. It sounds like he is trying to stretch the protection of cables (300.4) to boxes and misapplying 110.27(B)
A 2.5" deep box would not be compliant either then.
I doubt it would survive an appeal.

Ask him, "if the guy sistered op a short piece of 2x4 to the stud and nailed the box to that, would it be OK"?

That would give you 1.5" from the nearest edge of the framing member. It is not that I think that is what I would make them do but it might give you better insight into his thinking.
He may just not like plastic boxes. I have seen crusty old guys like that.

That sounds like a real scratch and stretch of 110.27(B).

Yes, technically, there will be 'live parts' within the box, but??

Although I cannot say a misplaced drywall screw entering a box has not happened; I have not heard of any.

That thought has never crossed my mind.

At one of the McCormick Place buildings in Chicago the EC got the brilliant idea to lay EMT in the troughs of the metal floor pan before the concrete was poured. It seemed quicker and cheaper than using chairs to suspend the conduits.

That worked out fine until the Carpenters started installing the suspended ceilings on the floor below...using power fasteners to attach their support wires to the deck.
They hit a LOT of the conduits that then had to be replaced with different conduit which now had to be installed overhead (20' ceilings) to power the receptacles and other loads.

Unfortunately for the EC the rework couldn't be claimed as an extra.

Let's flash back to a discussion we had about my 1957 house.

I had expressed amazement at the unfused, un-disconnect-able service wires that ran for about 20 feet, from the back of the meter pan to an interior panel, just under the surface of the bedroom drywall. Why, I observed, any errant nail could easily hit the wire, and you'd have no way to turn off the power.

Indeed, this very room had the drywall covered with cheap paneling ... long nails all over the place. It was a miracle they missed all the wires.

The result of our discussion? It was pointed out that cables /wires need protection ONLY where they penetrate studs; as soon as that wire leaves the stud, it needs no protection, no setback, from the wall face at all.

The same applies to the boxes mentioned by the OP. Unless the box is set into a notch in a stud, it needs no protection, no setback, no matter the material.

Hmmm 300.4(D)?

Many years back, a friend went to hang a picture in the condo. Got his drill to install a molly. He hit the adjacent condos 200 amp riser that was flopping in the wall cavity; not strapped to any studs. Ironic, it cost him over $2k to replace.

John...

At a brand new Albertsons grocery store, a trooper was using 4" flat head screws -- very high strength -- to anchor door stops so as to prevent wall damage at the handle height.

He found good purchase, he drilled deep into the bussing of a 200A three phase panel -- taking out two breakers. (Perfect aim)

The damage was mitigated because I always test for bolted shorts before powering up alien work. (stuff I have not installed myself.)

Upon first inspection, I thought that some conducting material had fallen behind the bussing and caused the short. My attempt to shift the bussing -- pulling it forward -- was for naught.

It was at that point I walked to the other side of the wall. BTW, backing that honking screw out really put a strain on my drill motor.

Even to this day I'm puzzled as to his expectation. Was blocking installed all over the job to await his rubber bumper installation?