I think that seeing offensiveness as a 2-place concept, with the elaboration that different sorts of judgments are more narrowly or broadly shared, or more or less important to enforce, is a cleaner distinction.

(For offensive things in particular, it also allows an interesting test: if Alice says something Bob finds offensive, one can ask Alice not if they think it is 'offensive,' but whether they thought it was 'Bob-offensive' before they said it. This helps separate out intent, and fix ignorance when possible.)