I thought Bill Barnwell had a measured take on it:
http://www.grantland.com/story/_/id/8389798/the-cardinals-forgot-lose-rest-notable-news-week-2-nfl
"I've seen suggestions that the Cardinals shouldn't have handed the ball off to Williams, but that seems like ex post facto logic. The Patriots had called their final timeout before that third down, with 1:10 remaining, but Williams had been effective running the ball during the final drive, and there was reason to believe that he could pick up a few extra yards and help produce better field position for the coming punt to New England. Williams is basically a rookie after missing the 2011 season with a knee injury, but research I've done has found that there's no significant difference in the fumble rates of rookies as opposed to more experienced players. And what were the Cardinals supposed to do instead, hand the ball to Beanie Wells (who has fumbled once every 63 touches as a pro) or let Kevin Kolb throw the ball? It was an acceptable risk that didn't work out."
Admittedly the pitch makes me nervous, but running the ball doesn't in this situation. The odds of a fumble on a given running play are small. The odds don't increase because it's near the end of the game. And if we are to believe Barnwell above, the fact that RW is a rookie may not be the factor some are suggesting.
The goal of the run is to reduce both time remaining and the Patriots' field position. We can't ignore the variables of the punt return, either. But you begin to turn them in your favor by increasing the ground the return must cover while reducing the time remaining after the return is over. Then Brady likely has one less play to work with, at a greater distance from field goal range. In Foxboro.
You're weighing the odds of a fumble against the odds of Brady and the Patriots offense getting the production they need with an extra play. They don't have that play if you deny it.
This is the logic I've seen behind the play call, and while this is a fair debate (as well as a fun one), I don't see how the running play can be flatly dismissed. The upside of executing a simple running play in this situation is equal to a critical stop on defense if it denies a play.