Capacitors, when they are "empty" can draw a lot of current very quickly when plugged-in. In fact, I bet some large capacitors somewhere is what is causing the arching. The arching is the least of your worries in such cases. It is the large current that flows after you plug-in that can damage under-specified electronics components that are in series with the capacitors. You need an in-rush current limiter to avoid that. Good electronic devices have that. I have a "manual" in-rush current limiter at home, for the rare occasion I disconnect my electric scooter's battery for maintenance - when it is reconnected it may damage the motor controller (not the charger). An incandescent light bulb in series between the battery and controller does the trick as it limits the current that flows through it - it lights-up for a second, then quickly dims down. At that point the capacitors in the motor controller are charged and can keep that charge for some time, so when the battery is connected there is no additional current, no sparks, no damage.
I'm sure something like that can be rigged-up to see what the Zeros' chargers are doing. My charger on the scooter does not start blasting at 15A the second it is plugged-in. It starts at milliamps and gradually ramps-up from there to whatever max current I have programmed in it. It does spark if I unplug it mid-charge at full power. But that is just a spark, not an increase in current like I described above, so no danger to electronic components. Does the Zero start drawing full Amps as soon as it is plugged-in? Does it have some beefy capacitors in the charger that get empty while the bike is unplugged and when plugged they draw way over the rated amps for a split second, potentially causing other in-series components to fail? I'd use an in-rush limiter to test that theory - the light should dim down after it first lights-up, provided the charger does not start blasting fill power immediately...
Now, don't you all go and electrocute yourselves trying it! Oobflyer probably knows exactly what I am talking about and likely still has his bulb-based inrush current limiter from his Vectrix. He can try it for us
I always thought electrical circuits had capacitors and suchlike to dampen these types of spike.