How does this generalize to firms with more than one stakeholder/owner? I don't see how it does without some magic where we assume that all members of e.g. the C-suite have similar, model-able reasoning.