[Orion.DeYoe]

We prototyped a belt shooter. We used 6 in. wheels for the rollers and direct drove it from a CIM. We used the orange urethane belt (just like the polycord everyone used last year, except flat) for the belt.
As soon as we powered it up we noticed a huge problem. When you get the CIM up to full speed (the kind of speed you would need to launch a frisbee a reasonable distance) the centrifugal force caused the belt to "float" off of the rollers.
We did some more work on the prototype and put in a ridiculously tight tensioner to keep the belt on the rollers. Then we ran the launcher at about 60% power. We were able to get 20-25 feet of distance launching at ~30 degrees.
This distance could be enough for some applications, but not for others. We're planning to prototype a two-wheeled shooter next.
We figured that if we got some belts that wouldn't stretch then we could use those, but our options seem to be pretty limited as to the belts we can get.
My conclusion? Yes, belts can be made to work, but you could waste a lot of time trying to get it to work consistently. Time that might be better spent on designing and building a good climbing mechanism.