Right now, a URL that's not recognized as belonging to one of the supported providers is returned as is. It would be great to return a snapshot of the web page instead, with some metadata that's retrieved form the page (e.g. title, meta tags, etc.)