That's the article PC was quoting...I think he even dropped a "knowledge is power" from that article on Saturday. But then again, that article doesn't speak to Saturday's situation because there was so much less time left in the Iowa game than the broader spectrum of the analysis on that page. While the 2 point conversion percentages/odds don't suggest a big difference, where the big difference comes into play is if you miss (as PC has said) the 2 point conversion the game is basically over at that point. That's why I felt it relevant to "normalize" the discussion a few posts up.
I saw Chip Kelly referenced earlier in this thread. As was mentioned, he used to take some of those gambles on 2 point conversions early (like, 1st quarter early) in the game, but often times he'd do it out of unique formations. Remember our lonesome polecat from the OSU game? Chip would run that and if the defense didn't adjust with enough bodies to account for the 6 OL and 1 RB-type way out to the left, he'd quick snap it to a QB/holder/kicker and swing it out to that back (standing behind the wall of OL) and basically tell that guy not to trip on his way into the end zone for the 2 point conversion.