The virtual Zendrum that somebody already did in flash Moon Rabbit mentioned above while I was typing this message would be a good app. What I’d really love would be every single word from the JoCo Looks Back vocal tracks, so I could make him say random stuff. Or every phoneme, so I could make him say even randomer stuff.
…actually… I could probably do that, hijacking parts of the existing Mac text-to-speech API to essentially make a JoCo text-to-speech voice. I don’t know how it would sound, probably extremely weird since it wouldn’t be a real MacInTalk voice and wouldn’t adapt the pitch appropriately or run one phoneme into another or do anything else that the actual text-to-speech engine does, but I know exactly how to do it (assuming JoCo has uttered every relevant phoneme at some point, at roughly the same pitch), and I think I’m going to have to try now. This would be even weirder than my robot choir.