If I understand your question correctly, our solution is dependent on the ICS. With ICS, each bottom bracket spindle is only handling one rider, so each Quarq is only measuring one rider. All measurement is done before the power is summed at the ICS intermediate driveshaft.
The cranks were not modified whatsoever (except perhaps for some accidental cuts during "engineering"). The Quarq center was cut large enough to fit around the "square taper section". Our dealer chose to minimize any "fastener torque", and therefore chose to sandwich the Quarq between the crankarm and an inner fastener. This inner fastener started life as the DV 34T ring that came off the arm, which was then cut down to a diameter that would cover the screw holes without interfering with the Quarq itself.
Because the DV 34T ring is "thin" and the Quarq is intended to have a 50/34 ringset ("thick"), the 34T-only Quarq is inset, and therefore presents a problem given how short the stoker chain is. A longer-spindle BB resolved that problem; since the Rear Admiral is already riding with pedal extensions, the BB change was perhaps a benefit. Some other BB "misadventures" led to an unfortunate scrape of paint on the R chainstay.
I don't have any pictures of the inner mechanics at the moment. The finished units can be seen in a pair of pictures on our dealer's website, houseoftandems.com (scroll down a little more than a screen's worth). If you'd like other views, let me know and I'll work on those next week.
For the moment, I'm using a Garmin Edge 705 and she has an Edge 800. I'm hoping to add two more Edges so that each of us can see both riders' efforts (I want to help her train smarter, as that was much of the justification for the project, and I'd like her help managing my efforts when the going gets tough in "the peloton").