Keeping it would mean every possible site parser in future needs to have it's own field.

Additional MDE plugins using other sites than IMDB or TMDB should use their own aspects/cache xml files and base them on imdb id.