Is numbers of papers published a good metric for determining the contribution of a company to public research?