It's really short: 1h10min. And yet, the camcoder conceit's long worn out its welcome by the time the first "thing" happens, at seventeen minutes in. Because, while you can allow set-up time for characters, situations, etc. in a "normal" film, you can't do that here: with the camcorder viewpoint, you're continually being made aware of Look, this is some character shit before the main attraction. Never mind the insanity of still carrying a camera around and pointing it at things, once everything kicks off.
There's some interesting stuff here: how the script is arranged to give the viewpoint character an excuse to look at things or be facing in particular directions, in order to catch them on film. But that doesn't hold interest, and I'd have preferred them to spend five minutes setting up a scenario where the viewpoint character is wearing the camera instead of carrying it. Because, honest to god, when you've got New York collapsing around you in massive chunks, are you really going to go back for a camcorder you've dropped on the floor?