I won't bother dazzling everyone with a list of qualifications, but you're really misrepresenting Positivism here. Positivism is really the idea that being empirically testable is a necessary condition for being both non-trivial and meaningful. Not (except maybe when being polemical, as Ayer could sometimes) that metaphysics is necessarily unverifiable.

Most of the problems we class as "metaphysics" probably aren't empirically verifiable, but if a few turn out to be, it isn't a big problem for the positivist program. Their objection to metaphysics isn't as principled as you make it out to be. It just falls out of verificationist semantics.

So if a physicist manages to empirically verify a few things that might have seemend metaphysical, then cool, turns out they were doing physics not metaphysics.