Because you can just extract any still frame from the entire video so you don't have to retry taking shot after shot trying to get it perfect. With video it has much more pictures to extract from the video because a few seconds already has a lot of videos? Or is picture quality from extraction from video lower than just taking a normal photo directly?