The notion that there is information to be gained by categorizing things after they are fully described is useless from a utilitarian perspective.

Thanks for this short phrasing for something I often want to say.

For example, if we know exactly what the process of waterboarding is, and how unpleasant it is, the answer to the question "Is waterboarding really torture," tells us nothing about the morality of doing it. At least that question might have some relevance when posed to presidential candidates, since "torture" is a legal category an saying "Yes, it is torture," might imply an obligation to prosecute waterboarders.

I agree with your connotation etc. - but I think the question "Is waterboarding really torture?" does have moral implications beyond presidential candidates: whether or not it is torture can determine whether or not waterboarding goes against a preexisting law or even informal promise ("No ma, I promise I won't torture anybody in Iraq"), and breach of agreement is morally relevant.

More generally, categorizing things even after they are fully described can still be a gain of information if the category label is mentioned in some outside agreement.

For another example, if Professor Witkins the Mineralogist told you "I'll give you $10 for each blegg you bring back from the mine, but nothing for rube.", and you're considering whether to put a purplish weird-shaped rock in your bag, even if you have full information on it you might still wonder if a Mineralogist would classify it as a blegg or a rube (Even if you know Witkins wants the bleggs for their vanadium, you still expect him to pay you for vanadiumless bleggs).