Ok technically you're correct, but pooling shouldn't be significant. The money you put into the pool is the collateral so having to maintain it individually or not should not be a major issue. Maintaining collateral for the shorts is just an expected part of the trade. I will concede that being long is more hassle-free, but I believe it's a minor difference.
That's like saying the risk of the entire system getting black swanned == the risk of one user getting margin called.
No that's not what I'm saying. I understand the thinking behind the design to pool collateral and use the least collateralized position as the trigger. It's an efficient design and the 200% collateral requirement will hold.
The risk of a short or long are understood before any trade: A black swan event happens when the least collateralized position loses more than 50% before it's forced out. Hence everyone should already account for an extremely improbable 50% drawdown event from a single trader in the system.
If traders understand that a 50% drawdown on any trader is systemic and both sides accept systemic risk each will go forward with trades. Short traders should not be concerned nor do I think they are concerned. A systemic factor like a black swan event you bring up effects both sides so it should have very little effect on the premium. Hence systemic risk factors should be separated from individual trading risk factors when evaluating the protocol. Then you'd be closer to identifying why premiums occur.