Appropriately weighting evidence, especially from public contributors

Robust handling of public contributions has not been tested against adversarial agents, and much of the contributor-facing infrastructure has not been built — user accounts, for example. No plan survives first contact with the enemy, but the same principles that allow Wikipedia to function should apply here too. Contributions will reach claim stewards only through intermediation with public-facing reception and routing agents, which will be instructed to be wary of manipulation. Abusive users and trolls will have their contributions flagged, and can be banned. Good reputations could be rewarded monetarily, as mentioned here.