Interesting concept, but what gives readers any authority to vote on the truthfulness of a story?
Perhaps a system where readers could 'endorse' or 'reject' a story and if a rejection threshold is reached it is evaluated double blind by a randomly selected pool of journalists with an accreditation score higher than some threshold (perhaps adjustable by the accreditation score of the challenged journalist as well). If a simple majority of the reviewers uphold the rejection it gets noted, and perhaps the accreditation score gets reduced of the journalist reporting poorly.
I'd have to run some simulations on that to see how whether that would be workable or result in overworked 'reviewing' journalists just rejecting or accepting everything enmasse.