Thank your for that idea. It sparked some of my own.
I guess it works as a 2D analogy of a closed universe, but it doesn't help jaketodd, since inhabitants on the balloon surface cannot experimentally determine the direction of the curvature (positive if on the outside, negative if on the inside, but this is impossible for the 2-dimensional inhabitants to determine).
Nevertheless, the balloon analogy is exellent for demonstrating that asking in what direction spacetime curves is nonsense. We can see that on the balloon, spacetime is embedded in 4 dimensional space (2 spatial dimensions, 1 temporal dimension and a fourth dimension into which spacetime also curves). By analogy we can see that we would need a 5-dimensional space in which to embed our 4-dimensional spacetime for us to be able to ask in which direction spacetime curves, and even then it would be a question of definition.