As Adams and Aizawa (2008), e.g., have suggested, cognitive processes trade in non-derived content and a couple system of the sort Clark and Chalmers (1998) envision could never be characterized in this manner.Now, this is technically not the A&A view. We do not maintain that a coupled system could never be said to have non-derived content. It's just that, as a matter of contingent empirical fact, this does not typically happen. We are not claiming that extracranial non-derived content or extended cognition is impossible.
But, I don't see that this technicality is crucial to Merritt's overall argument.