Unless I'm missing it, nowhere in the article or elsewhere did they say that ads only appeared one time.
They said that ads were served for one particular account 50 times (and presumably have data to back that up but I'm not inclined to give them benefit of the doubt). And that media matters had scrolled/refreshed a bunch of times to see whose ads would be displayed. Which seems reasonable to me.
Then TwitX made some claim about "50 out of 50 billion ads served" or something, which is a disingenuous comparison. This was one example of a problem. No one claimed it was the only example, so why would anyone compare against all ads served anywhere?