What doesn't ring true about this article is that the advertiser would spend $100K/day without building up to it with a whole lot of testing first.

Facebook may have declined requests for comment, but we are still only getting one side of the story, and I cannot completely believe this one side. There's more to it.