Sorry, I can't answer the poll until after the robot is shipped due to confidentiality agreements.
I recall four distinct proposals made by the first-week design subgroups. Two were for a straightforward one-wheel pitcher. One was a catapult. One was a monster spinning cylinder that could fling multiple balls at once. Of the prototypes, the catapult's aim was the most repeatable, but it was also the slowest by far, and there was no obvious mechanism for reloading it. Though there were no working pitcher prototypes, a softball-shooter derivative seemed the easiest to implement. When it came time to decide what we were going to build, even the person who designed the catapult mechanism voted against it.
(Note that I have fulfilled my commitment to secrecy, and haven't actually revealed what we ended up choosing.)