Simulation vs duplication seems like a syntactic distinction. If the simulation is faithfully duplicating the dynamics then the two are isomorphic, and that's sufficient to say they are, up to isomorphism, the same.
I guess one interesting experiment here would be if we had a computer running a biological simulation, and a computer running an LLM and they both had identical outputs given inputs, would you consider one of them more conscious?
Ooh, interesting. Another one: if we had a computer that ran a faithful replication of a conscious biological system, would you consider that conscious, even though it was running on silicon?