The problem is, even if you could develop a moderation system which can give individual reports this level of consideration, I don't think there's anyway to remove this kind of content without also removing a ton of legitimate discussions that happen to be used as dogwhistles in some specific subculture.