That flip side, where papers can be retracted when there is ideological opposition, was foremost in my mind when reading this! The accreditation algorithm can be tuned and weighted by user preference, or it should be. It is useful for those who do not understand a subject to have someone who does to review content in a way that , if you were inclined to understand it, you could learn.

Anonymous contributions could theoretically be submitted by a dedicated account. The reputation of that account would be subject to flux, though. A solution would be for that account to distribute the submissions to sub-accounts according to its own confidence. Only the reputation of the sub-account would be at stake.

We already do this with publications. If you say that the NYT has a history of quoting anonymous sources that turn out to be wrong then you have assessed their reputation. This will weigh on those who respect your judgement for the next time they include an anonymous source. The sub-account analog is found when people state their confidence in individual journalists.