When considering two theories that are qualitatively contradictory but have identical empirical implications, how can one say that one theory is particularly more valid than the other?
Heuristic principles (Least Destruction, Occam's Razor, etc) give us excellent critical tools for determining which theory to accept as the basis for further investigation (resting on an assumption of evolutionary progress towards greater knowledge, which seems valid enough to me) when we're dealing with tightly-bound and empirically-driven theories (theories of gravity, for example). These principles are less useful, however, when it is difficult to tell which theory best upholds them, such as may be the case with more holistic and wide-reaching metaphysical theories (theories of meaning and identity seem to be particularly obvious candidates for the non-applicability of Occam's Razor).
That having been said, selection between competing theories can become difficult. Do we accept pluralism, or do we find some other method of pluralistic defeat? Do we begin with the assumption of a singular truth, or may we allow qualitative contradictions so long as they are useful for our interpretations of reality?