It depends on what your implicit model of policing is. I think the real issue is that most police departments in the US are commonly dual-tasked with providing

It depends on what your implicit model of policing is. I think the real issue is that most police departments in the US are commonly dual-tasked with providing violent crime prevention _and_ front line social services. In this setting, it's less clear which "side of the force" is most impacted by the scientific input, but I think it might be reckless to simply cut them off without finding a better model for providing some of their more critical social services. In particular drug, mental health and domestic abuse interventions which sadly seems to be a very large portion of their workload.