There is one positive item, one huge negative and one poor implementation point.
The positive is that I think the mechanism, or put another way the Rosetta Stone platform, is mostly good. The combination of visual and aural, with a relatively simple screen-is-one-concept approach, is good.
The negative for me was the actual content itself. I learned about ladders, and horses and crazy things that not once have I encountered here in Paris. It would have been helpful to learn food items, grocery store items, things folks (even tourists) would likely find useful.
I hesitate to mention this because it might not have been a Rosetta Stone objective, nor might it fit into their framework/platform. But, there were no verb tense lessons. Even just rote learning of "je voudrais ..." [I would like ...] would get you out of 80% of the conversational jams when trying to get a sandwich, or similar.
The poor implementation point is that I did not experience any good luck using their voice recognition engine. Perhaps it is because I really have horrible French pronunciation. However, isn't that the very situation Rosetta is trying to correct? I found no way to discover why it was upset with my pronunciation. It was just a very frustration experience.