Evolution is theoretically, or analytically falsifiable within it's own argument, but it's premises themselves are not falsifiable.
That the external world exists is not falsifiable. In reality what you are doing is merely protecting this group from its own epistemic requirements of falsefiability rather than dividing what is or is not falsifiable. Falsifiability is thus just a heuristic, not an epistemology. Even so, very little of science is falsifiable even under it's self protection, much is overdetermined, much is active hypothesis, and ad hoc acquisition does much to barricade falsifiability.
TLDR - We should request falsifiability where applicable, but if we demand it for knowledge we lose nearly everything.