Copied from my post in another thread:
Originally Posted by
Robert C
First off, I have a NuVinci 360 on my commuter bike and with certain reservations I recommend it strongly. However, I will share this chart for those interested:
From and article translated here.
I think that the NuVinci did poorly because it really dos not fit the parameters of the test as well as the other IGH systems. Also note, there is no comparison of conventional derailleur based systems. Also, it should be noted that the graph starts at over 70% efficiency.
Even though I like the N360 on my Novara Gotham, I am thinking that I may start saving for a speedhub.