The service that is 800 feet from the transformer will have a difference in potential between the earth at the service location and the grounded conductor - unless there is no load on the grounded conductor. (it may still have voltage coming from the primary MGN though because of voltage drop on the MGN) Adding an electrode at/near this service will equalize that potential - to what degree really does depend on the resistance of the electrode and soil conditions. That doesn't mean you cannot exceed NEC minimum requirements and install a complex grounding electrode system, or in instances like we have with art 547 or 680 maybe even are required to install equipotential bonding grids or bond to additional items not required in general to ensure there are no voltage gradients in a specific area.
But this begs the question, should electricians pay so the POCO can save money?
What will keep the neutral at a light socket near the same potential as the plumbing or concrete slab is bonding within the structure. Granted a ground rod is another path back to the source lowering the impedance of the neutral circuit as whole so voltage drop is alleviated a bit, but in theory at the other end of the house if I took a volt meter to a spigot with the other lead in the dirt I would still measure voltage, voltage that would be much higher than bonded points within the structure.
In theory, if the earth was a copper ball soil would provide profound equal potential for miles around a single ground rod, but because soil is unreliable equal potential only exists directly around the ground rod if not right on it.
Intentional grounding creates a known reference point throughout the entire grid, the more intentional grounded points there are the more reliable that reference point becomes. Earth is a solid reference - but connecting to it with low resistance is not easy to do with what is really a small electrode surface area in general - that is part of why you have electrodes of some sort at nearly every structure connected to the power grid, including the service ends at each user.
But in theory, this is a waste of money. Would it not be more economical to just upsize or insulate the neutral than keep finding ways to turn the earth into a conductor? FWIW, Keep in mind that per NESC the neutral is only required to be earthed 4 times a mile.
I do agree that having an electrode present at all times or at all costs is not as critical as some maybe like to make it seem.
IMO, outside of a lightning strike near by I do not believe an earth rod does much.
I had an inspector once just about come unglued on a project that was not yet finished where I had not yet completed connection of the grounding electrode. He seemed to think that was very important or someone would be seriously be injured or killed, when my thoughts were that the POCO had an electrode at their transformer pole only 30 feet or so away, this was a non public building, and all concrete - so it was well earthed anyway, and I would have felt much more safe standing in that building during a lightning storm then outdoors with or without a GES, so yes I will install proper electrode when it is convenient for me and before calling my work completed, but was certainly not worried about not having this electrode the least bit in the meantime.
If you have a UFER the ground rods can be ditched. Id say the inspector is locked in the old school way of thinking that grounding is everything. Up until 30 years ago most books would actually depict ground rods clearing a fault, so the myth ingrained itself deeply.
How about one time recently when a local POCO was making a repair to an underground 34.5 KVA line. I don't know a lot about specific items or practices with this high of voltage, but electrical theory is still electrical theory. First thing is this was an old line that since has been replaced, but it did have a failure just a couple years ago and they needed to repair it and get it back on line. It ran adjacent to a barbed wire fence - like many lines in this rural area do. They had made their repair and still had the hole open when they energized the line. Apparently there was insulation breakdown somewhere in their splice - whether it was defective components or improper installation IDK, but they had an arc to that fence, and small grass fires here and there along that fence and arcing over to telephone pedestals near the fence for a few miles in any direction from the repair site, and telephone equipment damages in many customer locations in a few mile area vicinity of the repair site. At least they were smart enough to get everyone out and away from the hole before energizing the line. This is a good example IMO of how grounding helps protect from contact with higher voltage lines, had that arced to a well grounded distribution line or secondary line instead of the not so well grounded barbed wire fence - it may not have been as big of an incident damage wise as it would have cleared fault protection devices quicker then it did, and this location was within only a mile of the substation that supplied the line as well. Once it got much further then that distance away it became an overhead line.
If the line had all loads connected in delta (I assume it did not as is typical for North American) the fence would have cleared the fault immediately. Which brings up another situation, typical practices force high zero sequence pick up values. In order to trip high Z0 relaying a low impedance path must be developed. This in itself adds cost, complexity, and drastically increases the incident energy of a fault which also increase the duration to which voltage drop will occur across the grounding system ironically forcing greater equal potential boning and grounding.
I have seen a phase fall across an MGN some distance from the substation. The fault took a few seconds to clear the recloser, but pole pigs around the fault connected to the healthy phases experienced a significant voltage rise due to neutral shift in the MGN. In a case like this any other path back to the substation will reduce the shift mitigating the risk of customer damage due to over voltage. From this stand point extra ground rods do help, but mask what is IMO flawed engineering to start with.
However, I do agree with you in one case, if a phase fell onto a broken service drop having the ground rod might clear a protective device faster.
