No, it doesn't have to be all-or-nothing. The flaw in the Graham's argument is that he argues we should prefer "natural" without first proving "natural" means "good". Bubonic plague is very natural, but not so good, right? That doesn't have to be all or nothing, but each time somebody says "do X not Y because X is more natural", we ought to ask "but does natural here mean good?". Sometimes it does - eating an apple is probably better than ingesting a bunch of high-inducing refined sugar - but sometimes it doesn't. Antibiotics are not natural, but preferable to being eaten alive by bacteria. So this needs to be established on case by case basis - is "natural" good in this case? And Graham doesn't seem to bother to do that.