One last thing you might want to do with that o'scope just for grins. Look at the ac waveform when measuring between line and neutral. I suspect that it will be cleaner than between line and ground.Hopefully this adventure will be helpful to others as well. It's educational for me, for sure.
I got the bulb hooked up and it doesn't light up. There's some voltage reading across it, the volt meter jiggles around a bit but stays under 5. So it's not enough to light anything up.
With the bulb plugged in, line to ground reads 124 or so.
Without the bulb plugged, the ESP surge protector won't transmit current. With the bulb plugged in, it does. Presumably, the voltage on the neutral triggers something in the ESP and shuts it down.
So bonding the neutral does indeed straighten out the signal generation. The oscilloscope shows a much better looking signal than before. And it's filterable. Running it through the ESP gives a perfect looking signal that's actually better than the utility mains.
So it's probably very possible to set up two generators as the OP originally suggested, as long as the wiring and grounding can handle the completely unbalanced load.
My last question would be in the construction of a grounding plug. Is it better to wire up a dead short between the neutral and ground to bond the generator, or better to wire in a resistor in place of the incandescent bulb?
RE your last question: The reason I initially suggested testing with the light bulb was to limit the current from neutral to ground in case the Predator's output wasn't floating. It's clear now that it is. Neutral/ground bonding is supposed to be low impedance (low resistance) so a dead short is the norm.
BTW, if you tie the generator into the house wiring as the OP intends to do, the neutral ground bonding will occur within the house wiring, and no additional bonding will be necessary.