As an aside, many people have trouble understanding that a lot of legal proceedings are about meta-level issues (whether and how a case can proceed, as opposed to the merits).
I'm not sure I agree. Many legal doctrines, especially procedural doctrines (like jurisdiction), are justified based on knock-on effects. But that is different than recursive analysis.
Or, to use LW terminology, meta != decision-theoretic reasoning.
Some legal background:
One would think that disagreement between Circuits about the meaning of a law would be legally relevant evidence about whether the law was ambiguous. Instead, there appears to be a circuit split on the meaning of circuit splits.
More available here, for the amusement of those on this site who like to think meta. Also a bit of a lesson on the limits of meta-style analysis in solving actual problems.