I would also classify myself as a 'semi-theorist' ... I was speaking with a theory professor at a university that offers a subfield in methods, and she told me that she recommends all of her theory PhD students sit their second comp in methods, unless there's a very compelling reason to sit a different one. Some people who are more semi-theorists, like myself, end up doing things like comparative for their second comp, but she recommends people who do pure theory to always choose methods for their second comp.
I have to admit, I was surprised by her suggestion, but when we discussed the matter it made perfect sense. Methods makes you more employable, and theory is the field that makes you least employable. Also, doing a research methods subfield requires a depth of understanding of ontological and epistemological frameworks beyond that of any other subfield, making it relevant to theory.
My university doesn't offer a methods subfield, but we're all encouraged to do methods courses. I don't know if I'll ever use them again, but I'm a better theorist for having taken methods courses.