Years ago it used to be that the ampacity tables for UL508A were different from the NEC tables (allowing more amps), which made for a lot of tense phone calls in the field. Now UL has aligned themselves with the NEC for what panel shops can do, but that does not apply to a UL listed DEVICE itself. The mfr can use whatever they think will pass the UL specifications, which include temperature rise inside of the enclosure at whatever current it is rated for. So if 1/0 does not cause the internal air temperature of the switch to go above 75C when 200A is flowing in a 40C ambient, then it passes.
If you remove that wire, then decide later that you want to put one back in, YOU cannot use 1/0 again.
When switching DC, the arc will last longer because unlike AC where the voltage crosses zero 120 times per second, DC stays steady and even as the contacts start to open, the ionized air helps to keep current flowing until the arc is too long for it to sustain itself. One way to do it is a very long disconnect switch distance so that the arc length is increased quickly as the switch opens. But that's impractical in most cases because the switch itself becomes long. But since it is the total arc length that counts, another way is to split the arc between multiple sets of contacts, allowing the size of the switch body itself to be smaller. In this case, each set of contacts is 'double break", so from the total arc length standpoint, two sets in series gets you 4 times the arc length and the switch itself can be more compact.
It's in
here, page 19.
PS:
Oh damn, I forgot to look at the number of pages again. I'm late to the party on this....