it z-talks about consciousness for the same reason (at the appropriate level of abstraction) as the conscious monad

"At the appropriate level of abstraction" is pretty broad. Part of the reason that a conscious monad talks about seeing colors is because it does see colors, whereas its simulation (let us suppose) talks about seeing colors only because it contains computational tokens imitating the causal role that colors play in the conscious monad's internal transitions of state. I don't see any contradiction. How would you employ a standard anti-zombie argument here?