I'm probably not thinking this through, so go easy, but why wouldn't tracked panels also produce a normalized power curve? My assumption behind the question is that they follow the same track as Sol every day, which doesn't vary at least year-to-year.