If something is true, it ought to make a difference, and in some sense that difference ought to be verifiable. Generically, this is the space inhabited by logical positivism, but the logical positivists had rather specific, foundationalist notions of verifiability that I would not wish to follow. (Moritz Schlick, the founder of the Vienna Circle, spoke passionately of verification against “the Given”, which was supposed to be a bedrock of pure, uninterpreted empirical fact that would anchor the whole enterprise of science. He also literally talked about foundational “pointing”. But he had a good critique of epistemic claims for intuition and images; emphasized conceptual development, form, and structure; made interesting use of relations; and reportedly spoke of laws of nature as inference rules.)
Logic by itself will not reform the world. However, the analysis of illogic is generally salutary.
The kind of verification that seems most applicable to the sorts of meta-ethical theses I am mainly interested in would be pragmatic. I imagine general pragmatic verifiability as just extensive openness to rational examination, with a responsibility for due diligence. Obviously, this is a loose criterion, but as Aristotle would remind us, we should not seek more precision than is appropriate to the subject matter.
In principle, material-inferential things can be verified “as far as you like” by a sort of recursive expansion of material consequences and material incompatibilities.
Purely formal-inferential things can be rigorously verified by mathematical construction or something resembling it. In constructive logic, proof comes before truth, so verifiability is built in.