I actually don't think the big hurdle would be the business logic of individual systems. I think the big hurdle would be forking Pleroma, comprehending the codebase enough to modify it to include this new thing in it, and then structing that part of the code to coherently allow for freeform interpretation_system, well, interpretation. You'd basically have to write a little framework into it outright, much like MRFs.