Subsidiarity and Religious Establishments

This paper, by Kyle Duncan, "Subsidiarity and Religious Establishments in the U.S. Constitution," looks interesting.  Here is the SSRN abstract:

This article proposes subsidiarity as a tool for understanding the problem of religious establishments and the function of the Establishment Clause of the U.S. Constitution. Subsidiarity is a theory with roots deep in European political thought. It concerns how persons become genuinely free by associating with others, and what those associations imply about state authority. Subsidiarity’s goal is to empower assocations by conditioning the intervention of state authority. State authority should foster, protect, and coordinate associations, but never absorb their functions. In this way, associations will provide genuine freedom and development for the individuals within them. As a political theory, subsidiarity has been applied to a wide array of problems—from international human rights and the International Criminal Court, to corporate governance and the European Union. Subsidiarity promises interesting applications to establishments and the Establishment Clause. The article concludes that, as a substantive norm, subsidiarity provides a viewpoint from which to assess the role of religious associations within a pluralistic society. It also furnishes an illuminating way of understanding the problem of religious establishments as such. As a structural norm, subsidiarity helps place the Establishment Clause within the federal framework of the U.S. Constitution. It invites us to view the Clause as a structural strategy for dealing with the problem of religious establishments faced by the authors and ratifiers of the U.S. Constitution. The conclusions promise a helpful reorientation of the typical discourse about religious establishments in U.S. jurisprudence and scholarship.

As I have tried to explain elsewhere, I am a big fan of subsidiarity as a political norm and structural principle, and am also inclined to regard as important mediating associations and their freedoms.  That said, I wonder if there are limits to subsidiarity's usefulness when it comes to "[]orienting [our] discourse" about our Constitution?  Subsidiarity, as I understand it, is a principle according to which decisionmaking ought to take place at the "lowest" possible level.  But, subsidiarity is not a general requirement of decentralization or devolution; sometimes, when centralization works best (all things considered), then centralization is required.  Under our Constitution, it seems to me, the local v. national decision is often (not always, of course) a function not so much of "which level works best?", or "which level of decisionmaking best facilitates human flourishing?", but "is the government actor in question authorized to make the decision, or engage in the conduct, at issue?"  If I'm right about this, then what room is there, really, for "subsidiarity" in deciding Religion Clause cases?  (Read the article, Rick.  ed.)