Wow -- I've read some really well-considered responses, but this one might top them all -- thanks!!
As nicobie (OP) stated at the beginning of t(his) thread, I think this needs to be done in the US 240VAC single "split-phase" world, although I'm not positive about this. The Enphase microinverters are spec'ed to have their AC output on two wires, both 'hot' with 120V sine waves on each, 180 degrees out of phase, to get to the RMS 240VAC. In the event of multiple microinverters, they are simply paralled until they reach a certain theoretical maximum current flow (~ 10-16 microinverters) per array.
As it happens, my wife recently upgraded us to a new table saw, so I have available an older 10" saw motor I have wired for 240VAC. I don't yet have a 240VAC inverter, but would be perfectly happy to add that for educational purposes. I guess I'll have to do the math about current draw of the motor at no-load and match that to the inverter's rating, plus deciding what battery to use it with. This would form the proposed off-grid 'microgrid', to *fool* the microinverters, correct?
If I understand this setup correctly, the major concern would be to not overload the 'starter' inverter via back-feeding it, once the solar panels start pushing electrons through the microinverter -- which will turn on once they detect that the microgrid is up and running. Managing this power flow without blowing up the inverter is where I'm still fuzzy, I guess. My first step probably should be to find a bi-directional wattmeter for 240VAC circuits. BTW, I'm perfectly happy to 'blow shit up' (other than my shop or body) for the purpose of our collective educational advancement!