Maybe the problem set they are looking at, commercial riders on a schedule, is very different from the wide ranging meta study that you linked.