Via Brian Leiter (here), I discovered this strange blog post by Les Green, a philosophy of law professor at Oxford University. In summary, Professor Green offers a nuanced critique of the “pursuit of novelty” in fields like jurisprudence (philosophy of law): the pursuit of the new, which has now become a “fetish” (to quote Green), should take a back seat to the central goal of his field: the pursuit of truth. Professor Green’s diagnosis, however, has a fatal flaw: he fails to define truth. How do jurisprudes, for example, determine whether “the next general theory of law” — or even any of those listed in the picture below — is true or not? Alas, there is no way of testing, refuting, or “falsifying” (in the Popperian sense) such general legal theories; in a purely literary or linguistic field like jurisprudence, it’s all just a matter of opinion! As a result, although novelty is an imperfect proxy for truth (indeed, there might even be an inverse relation between novelty and truth), in a field like jurisprudence it is much easier to determine whether a theory is new than true.
prior probability 
Truth can be difficult to determine, so I agree legal theory provide us with useful tools for “approximating truth”. Similar to why most schools of economic thought (Austrian Economics, obviously being the most notable exception) uses economic modeling to approximate market conditions. The result will be static results do not take into account for the fluidity and dynamism of actual practice.
However, the better question is which school of legal thought gets us the approximation that most resembles the truth?
excellent observations; it generally comes down to “theory choice” — specifically, what criteria (parsimony, explanatory power, beauty, etc.) we should use to choose among competing theories
Maybe it would be best to use several effective theories to analysis an issue rather than adhering to only one theory.
I would also assume that “effective” theories would have some overlap; their performance would be indicative of conveying some degree of truth. This might also be faulty reasoning on my part.