Yes, we can!
Update: here's a BlenderArtists thread with a description of a similar setup.
This is based on another Blender mocap video test out there. I used Blender 2.61 with motion tracking to track and save the face marker points from the video and then plugged in my cartoon model. There were very few markers on the face, and it would have been better if there had been a few more. As it was I had to overlay the nose wrinkle and lip sneer along with helping out the enunciation of a couple of words. I could refine it further, but I think it would be better to start over with a better source video with more markers. Anyway, no-cost face capture! Blender rocks.
Some more tests: