A lot of this is intriguing, but I'm confused by the focus on preventing fraud through making it easier to deny claims -
We're incentivizing the community be on "the winning side" of the vote so, unless that bias exists collectively, the voting process will pick the right outcome in the long run. "Guaranteeing" a claim is a risky proposition as that would leave us open to fraud - that would also require a central authority to basically say "no, we override the community" which kind of kills the decentralized concept in the first place. The community will be able to have an open discussion on claims as they're voting to compile/share evidence. Worst case scenario as a policy holder is you get your money back.
Also, to be fair, anytime you're trying to extract money out of someone for a cause, you have to provide whatever they claim is sufficient evidence. The UMC community will be a reflection of all of our policyholders.
As a community member with a stake in the float, don't I actually just want to deny every single claim including the good ones just to keep my own stake higher? If anything I would expect there to be a sizable group of people who try to be as stingy as possible. Related to this, we all know that the (real or not) hot girl in trouble is going to get 3x the votes a man in the same situation would, another sizable chunk of the community will be pretty biased against everyone who doesn't look like them and so on. Decentralizing oracles work for prediction markets where there is a clear right answer that anybody can look up, but I don't see a good way to make this work for something with this much subjectivity. Am I missing something obvious? How do you plan on incentivizing the 'right' side as opposed to simply the side with the bigger bias?