The obvious middle ground is to allow publication under controlled conditions; i.e. declaring the research results a state secret and managing access.