I think compartmentalising can be useful.
Mises' methodological dualism is about using different methodologies in natural sciences and social sciences.
I'd say that the spritual realm and metaphysics can be approached differently from how we approach politics and economics as well.
Sure, in a pluralistic/classical liberal society, we don't have to agree _why_ we shouldn't lie, cheat, or kill--so long as we all agree that we shouldn't, then we can live peacably.
But when the question does turn to why--say, over a fine meal--we have to get 'back of' those assumptions. "Why do you believe that? How can you say that consistently? Etc." Many people can count very well, but they cannot _account_ for counting. Many can reason well (like Mises) but they cannot _account_ for the capacity to reason--where did it come from? What does it reveal? Etc. But when the plumber comes over, we don't discuss metaphsyics. He knows how to plumb and, for those purposes, that is enough.
Yep, I think that's beyond the scope of what Mises has written. He made a similar admission in the first part of Human Action, if I'm not wrong.
Imo, both rationalism and empiricism have their limits, meaning they are useful and fully valid in explaining the external world. When it comes to the world within ourselves, their limits become clear.
