So if metadata is somehow distinct and separate from what it describes, does this mean that the shutdown of sites pointing to copyright-infringing content will stop?