According to Wikipedia, a Hail Mary pass is a very long forward pass in American football, made in desperation with only a small chance of success. Seeing YouTube's numbers, you have to be reluctant to think of "desperation" and "small chance of success". Then again, you hear rumors about YouTube paying huge bonuses to their big channels to not abandon their platform as soon as Facebook is finished with suggested video. And Facebook announces they have 8 bn video views a day, doubled since April, and without even having really rolled ou the video news feed, the true product behind all this.
Who cares of FB counts three seconds of autoplay as a video view and that in minutes served, YouTube must still be far ahead of Facebook? How far, that is the question, and can they stop or reverse the dynamic? One step to do so could be the embedded player. News and video sites had to invest a lot of money in video display and delivery, content and asset management etc., just so that they could serve their own video pre- and midroll ads on their own video infrastructure. That time seems to be over soon since YouTube will offer their technology to them. As a side effect, their content is so easily published on YouTube, they won't abandon the site. Could be a hail Mary after all.