I’ve just come across yet another file format registry: the P2 Registry at the University of Southampton in the UK. It’s identified as a beta and was pretty slow when I tried it, but it has some interesting features, including risk assessments of formats. PRONOM and other data sources are used. There is a short PDF article on the aims of P2, which tells us that “the key feature of the registry is the ability to import arbitrary ontologies that can be used both to infer new facts from existing information as well as to align (in the case where two concepts are similar or the same in nature) information already in the registry.”
Its web user interface is minimal at the moment, but it’s worth keeping an eye on this.