Knowability and 'Valuability'

One reason the knowability paradox seems so paradoxical is that the principle of knowability (taken on the paradox to entail the omniscience principle) seems so darn plausible. That principle states that all truths are knowable. Suppose this modal principle were extended to evaluations. We might say then that if X has value, then it is valuable--that is, it is possible that someone, somewhere, at some time, value X. Jointly, the principle of knowability and the principle of valuability imply that if some X has value, then it's possible that one know that she values x. Problematically, though, if knowing p is more epistemically valuable than merely truly believing p, then by the principle of valuability, it is possible that one value knowledge of p to her own mere true belief that p; and by the principle of knowability, it is possible that one know that she values knowledge that p above her mere true belief that p. But given the fact that we can't know of any mere true belief we have 'that' it is a mere true belief, (i.e. given that (p & ~Kp) is not a knowable proposition), it seems as though we must reject that anyone could know THAT she values knowing p over her mere true belief that p. But given the principle of valuability, this would imply that knowledge that p isn't more valuable than mere true belief that p. But it is! It seems, then, that we've got a paradox of valuablility that runs parallel to the paradox of knowability.