We don't yet have a test to ensure that all strings are localized (i.e. don't match the default language). Difficulties in this one would be covering languages where translations are identical (en -> en-GB, etc.)