Jamie, it is a neat idea, I only see one point that causes some concern. (other than fabrication details, but it sounds like you are capabile in that area) To achieve effecient heat transfer from the cool intake, to the air you must have a high surface area for heat transfer, lots of turbulance, and a large temperature difference. I see a slight problem, in that ideally as far as flow rates are concerned, you'd like your intake to be smooth on the inside, and not to induce high amounted of turbulance. (which isn't always the case in a carb'd setup) I'm assuming for this trouble you are running a turbo/injection.
To achieve a high rate of heat transfer, you basically want the inside of your intake to look like a radiator. Lots of fins, lots of surface area, lots of turbulance most likely. However, to achieve a high rate of airflow, you want the exact opposite situation in your intake manifold.
It would be an interesting experiment. Off the top of my head I'd say you'd see more gain in low and mid-range power than top end. However, if you don't rough up your intake walls significantly, I'm not sure you'll see much effect at all as far as temperature drops in the intake charge.
My 2 cents. I'd love to see it attempted and hear about some results!