I would think that it’s still easier, given the technology, to do the performance capture rather than program in actions, emotions, and speech from scratch. So as long as the scene in question uses humans or anthropomorphic creatures I think you will continue to see motion capture dominate.