As of yesterday, despite the rocking of the story and game concept, I was still convinced that using a game engine and machinima was the best approach. I'd even presented various examples of the genre so people knew what I was talking about (only two or three had ever heard of machinima). But after seeing a very cool game built by MIT using 2D cartoons and animation, I started thinking. Then after discussion around the table, I have to admit that I swung 180 degrees. Going from 3D game engine to 2D with a classic multi-panel cartoon design solves a lot of issues wrt look and feel, balancing computational horsepower for speech recognition, and enabling amateur cultural production and involving students in helping to create and extend the stories.
Hey, I can change my mind. It's my perogative. My feminine side...