Right now in ClaimChaim we speak about equivocation by users. If we want to stick to your story then this needs to be changed, and only added as a clarification at the end.
An alternative, is to actually consider non-honest users in the attacker model. Does this actually break any of the other defenses?
@carmelatroncoso in #40 :
We need to think this through.