I'm not talking about direct use of fMRI, but rather that some mental health experts are so laser-focused on the problem of untreated mental illness (not without reason, it should be said) that they argue for widely deploying interventions merely because those interventions are supposed to work rather than because they're demonstrated to work.