I was going to mention micro-inverters, but I just started researching them and don't know enough yet to provide any useful input. Download the Owner's Manual which has much more info in it.
Magnum Dimensions MicroGT 500 Inverter
Um...no.
Most (if not all) micro-inverters are "grid-tie" inverters that sync phase and frequency to the grid. If they don't detect grid power to sync to, they shut down their AC output.
The manual for those Magnum Sensata units describes a way to fool them into working in the absence of grid power. You feed a normal inverter from batteries, and the micro-inverters sync to the output of the battery-fed inverter.
Not really practical for most mobile installations. Plus, how you gonna charge the batteries with that kind of setup?
There is something hinky about that Gone with the Wynns series vs. parallel test - but I haven't thought it through enough to spot the flaw.
Still...in series, shading one cell should not drop the output of the array of two panels to zero.
I'm pretty sure they goofed up the test somehow.
Edit: Could be that the Outback charge controller only does an MPPT sweep (adjusting array load voltage to determine Vmp) every sixty seconds. If so, they would have to shade a cell, then wait for the next sweep.
In parallel, shading wouldn't change the array voltage (array Vmp), so the MPPT wouldn't need to readjust the load voltage. In series, partial shading would change the array Vmp, so the MPPT would have to find the new Vmp.
So maybe they just didn't wait long enough for the MPPT to adjust...
Dunno. Prolly have to sleep on it.