I wouldn't expect that the efficiency loss would be enough that you could tell just riding the two back to back without power data.

I would expect however that its probable that the ICS system is costingabout 2% of power in efficiency loss.

Forum member ASU GT did testing with SRM cranks and a power tap hub, on a conventional tandem, to measure drive train loss, and found that power loss for the stoker was 1.8% and power loss for the captain was 6%. Other studies show driverain loss from a bike chain at 2-4% of power.

The Davinci adds another bottom bracket, 2 chainrings, freehub type mechanism, and chain. Conservatively, it would be reasonable to expect that the friction loss from all that would be in the range of the 1.8% power loss the stoker experiences on a conventional tandem. It could be in the 2-4% range found for drivetrain loss on singles.

There are additional reasons to believe that the loss could be higher. First, it's a known fact that smaller cogs have more friction loss than larger cogs. The tiny cogs the Davinci uses have to be adding some friction loss. Second, ASU Gt's study showed that the Captain's efficiency loss was greater than 6% if the stoker didn't pedal. Thus, to the extent, the ICS ends up with more time where the Captain pedals and the Stoker coasts, it increases the Captain's efficiency loss.

You could prove this out empirically with 2 sets of Vector pedals, and a power tap. Until someone does that we're limited to extrapolating from existing data.

Admittedly a 2% drivetrain loss would hardly be noticeable, but it is a tradeoff. Which is my whole point, the ICS system involves tradeoffs. For many those tradeoffs are worth it, for others not.

For us, we have no desire to independently coast, so any downside at all to ICS comes with on offsetting advantage. That's obviously not the case for anumbe of other who clearly love ICS.