Tests such as this often generate more questions than answers.
1) What was the distance from the camera to the subject, and from the subject to the wall behind her (this seems to vary somewhat, especially with the first shot)?
2) What was the color and height of the ceiling, and how high was the subject (standing or sitting)?
3) What zoom setting was used on the flash? This could presumably change how it bounces off the card/diffuser/ceiling.
4) Has anyone done similar comparisons with Gary Fong’s Lightsphere II?
To my untrained eye, the index card seems to be as good as the more expensive options. It would seem to be more efficient, doesn’t take up much storage space, and has other uses. Perhaps those more experienced in evaluating portrait lighting can see differences that I can’t.
Regarding your comment on the omnibounce: Even when pointed at the subject, it would still bounce some light off the ceiling/walls/etc. Outdoors would be a different story.