@adam
The ideal system IMO is not automated. It is threefold
1 - A solid set of easy to understand rules defining inappropriate behavior within our community
2 - Allowing people to bet if a given post/comment/whatever does or does not comply with the rules. There is no advantage to betting it does follow the rules unless you think other people are incorrectly identifying it as a rule-breaker when it is not.
3 - Have non-controversial ones be settled automatically
Settled as in no tokens change hands and the incentive is reddit-like content weeding?
4 - Have controversial ones flag a moderator who then does a rules-judgement, losing side donates their tokens to the winning side.
It does seem a sound approach, but I am reminded of an anecdote about a child daycare facility that, because parents often picked up their kids late, started charging for children staying over time. The response of the parents was to treat it as a service they could buy more of, which was not the intent of the administrators. Abolishing the fee didn't improve things either since the parents now valued the service monetarily and were getting it for zero marginal cost.
If moral/community obligation is mixed with monetary incentive, the latter could wreck the perception of the good that underlies the former. Just an observation that may have only limited applicability to this situation - I think it will be an interesting experiment.
Cheers