To me there seem to be many examples of good impactful strategy research that don't introduce big new ontologies or go via illegibility:
I do also see examples of big contributions that are in the form of new ontologies like reframing superintelligence. But these seem less common to me.