I'm not 100% clear on what you mean by the above, but...

To measure total power with a crank-based powermeter requires instrumenting either both sides (e.g., the original Rotor powermeter, Verve), or the spider (e.g., SRM, Quarq). If you instrument only the left side (e.g., Stages, Rotor LT) or the spindle (axle) (e.g., Ergomo, new Rotor), power supplied by the right leg isn't sensed, as it is transmitted directly to the drivetrain.

ETA: I don't recall seeing these sentences in the original article (although it is quite possible that I missed them):

"It can only measure from the left because the force coming from the left crankarm being pedaled has to travel through the spindle to get to the spider. On the driveside, your power is going directly to the spider, which is why so many crank based power meters measure power there. Those measure aggregate power without left or right information. Here, it’s doubling the left side’s number to give you total power."

What's really confusing, though, is the shot of the head unit showing pedal smoothness for both the left and right legs, which isn't possible. Perhaps the fact that the crank being tested was a prototype means that data were being misreported via the ANT+ protocol, thus accounting for this error as well as the fact that torque effectiveness wasn't reported (BTW, contrary to the article neither of those metrics are new, being developed/introduced in response to release of the original Rotor powermeter.)

Last comment: my right leg is particularly "smooth", but even it never reaches 35% (of course, there's also no evidence that maximizing "smoothness" improves performance/function in any way...in fact, all of the data point in the opposite direction).