It isn't much better for simple animations like this but I find it much easier for complex, gesture-based, dynamic animations that can change at any time based on user input.

Also a designer can tell a dev exactly what he/she wants in terms of animation values (assuming they're using origami)

Given that much of the Paper prototyping was done in Origami, this was my first thought when I saw this library. I haven't actually used Paper yet (it's still US-only) but I'm in love with these kinds of tactile, fun interfaces.

Makes sense.

I think the main difference is they are easy to interrupt and continue from the current real position, where doing that with CA is more difficult. That would let you do some nifty stuff with gestures, I guess.