Models and the modeled
A common pitfall is to design (formal) models for formalism’s sake. That is fine, and can lead to good mathematics.
However, when modeling defeasible argumentation the formalisms are not studied for themselves, but are meant to be about something, viz. about defeasible argumentation.
Models hang in the air if they are not accompanied by intuitive, extra-systematic accounts of the modeled.
(Cf., e.g., Susan Haack in Philosophy of logics)