1) the preemptive rejection of the strong naturalist intention (study ethics and politics solely in view of empirical evidence) does not offer any good reason why it should be so.
1.1) one good reason I would see is that, while we can understand evolutionary impacts of certain kinds of political and ethical arrangements (and under which conditions they come into being), effectively testing them would demand another framework – one such as federalism or the archipelago.
1.2) the reason “it assumes that empirical knowledge is superior to ethical and political knowledge” is weak. the naturalist method has produced demonstrable results for over 4 centuries now, incarnated in the technological world all around us. ethics and politics have produced nothing of the kind. if that’s not an evidence against the method used in ethics and politics so far (an introspective method, based at best on logic and at worst on suspicious “a priori” definitions).
thus it’s not that empirical knowledge is superior to ethics and politics, but that the empirical method works better than the introspective method used in ethics and politics.
2) the attempt to evade the facts of empirical knowledge (for whatever reason) does not make certain political or ethical theories any better adapted to empirical reality. their flaws and their course when realistically applied can be empirically predicted. refusing to listen and pay careful attention to what naturalist empirical knowledge has to say is undertaken at the critic’s own risk. reality rules.
3) there’s an universal impartial judge, and it’s called reality. reality shows itself through survival. irrealism will lead to death, necessarily. those that survive the most, have the most truth in their beliefs and practices. those that die, don’t.
3.1) sure, the contest may last a long while. and to this extent, two different and even contradictory sets of beliefs may be held as truth, at the same time. in such a point in time, both must be held as truth indeed. when one dies, then the other must be recognized as the truth that prevailed.
3.2) once again, choose to ignore reality at your own peril.
3.3) theories, of course, evolve. hopefully, learning from deaths around itself. if a theory that leads to longer survival is abandoned, for whatever reason, the whole of the institutions and societies that embed and accept them will eventually die, especially when in competition from societies that haven’t committed the same mistakes (and will themselves learn quite a bit about how not to run institutions and societies).
4) for normativity, see “The Blind Mechanic II”
5) Only naturalist methodology is “legitimate” because only it works as a good proxy for real consequences, and thus ensures realism. all other methods are greatly liable to delusion and willful blindness.