Salvo wrote:
The current plots tell the story.  The current pulse time for the Onan is almost twice as long as the Honda.  That means the cap bank voltage drops a lot lower when powered by the Onan.  Low cap bank voltage means converter goes out of regulation. 

Onan

Honda EU2000i
I don't agree that the "current plots tell the story" or that a longer current pulse shows any kind of problem.  For example, a perfect converter/charger would have a power factor (PF) of unity and would draw current as soon as voltage rose above zero.  An expensive PF corrected converter charger would have a sine wave shaped current pulse, producing the maximum possible width for each pulse. 
Of course, I'm not saying that the PD has power factor correction, nor that the PD and Onan MicroQuiet combination works as well or better than the EU.  It's just that the width of the current pulse alone doesn't seem to tell us much, other than that the charger is taking energy from the gen over a longer period of time - something that is usually good.  
The total energy extracted from the generator during each cycle is what we'd really like to know, since that's the energy the charger/converter is sending out its output (minus some efficiency losses).  To get that power number, we'd need to consider the shape of the voltage waveform, as well as the current waveform.