There is clearly some confusion here because the NEC and NFPA are saying that these devices draw more amperage to make up for low voltage, balancing the equation. And the only time the NFPA gets involved is when there is risk of fire - over-heated electrical in this case. But I guess I should say perceived or theoretical risk, because the NFPA does go a bit over the top.
But Hughes says: “The Autoformer DOES NOT take power from the park. It does not affect the park or input voltage, or make electricity. What it is doing is changing the voltage – amperage relationship lowering the amperage [tapping point] and raising the voltage.” By doing this it is also lowering the total available wattage. Something to keep in mind - you can‘t run as much stuff, but at least what you run will be at the proper voltage
The only way it can do this without drawing additional amperage from the pedestal is by lowering the available amperage to the coach (Hughes’ claim), robbing amps to pay volts. This makes sense (again balancing the equation) to protect what you are running from low voltage, but it also reduces your 50amp service, depending on the adjustment, on the coach side of the autoformer. I’d say this is ok, since, if the need exists, you are in a low power situation anyway. Better to regulate it on your side of the pedestal.
Like Mike Sokol, I’d be interested to find out which is true. The test would be simple to perform, but would also be expensive because you’d have to buy an auto-former. So this question will have to be answered by a disinterested third party in possession of an autoformer, and with the curiosity and wherewithal (power supply) to perform the test.
My guess is that Hughes is correct in this case, and the NEC and NFPA are addressing all auto-transformers generically without regard to the specifics of the Hughes device. Otherwise they would have to test all auto-transformers and then specify which ones meet the code and which ones don’t.