Weight matters a lot for bikes but this is sort of a solution to the wrong problem. A lot of the energy would likely get lost as this thing spins at high RPM loosely transferring some of its energy into forward momentum.
Because it relies on friction against the rim it seems like if you applies force on the wheel (like going up hill) it would just spin freely against the rim. Because the biggest contributor of weight in e-bikes is by far the battery I can't think of many scenarios where the weight this saves would be worth the efficiency loss that you would have to compensate with a much heavier/more expensive battery.
Maybe its just me and I don't follow e-bike design really but this doesn't seem that difficult to solve. Instead of applying force to the outer edge of the rim why not look at something like the front disk brakes as an example?
A disk/cog with teeth could be added to the right side of the front tire meant to pair with the electric drive system so you wouldn't see high system losses from slipping. You could engage/disengage it allowing the bike to freewheel similarly.

@Daishi
Because of the offset/angle of the mount the traction from the drive motor on the rim requires a counteracting moment. The motor looks like it is free to rotate so it is up to the normal force of the traction drive wheel against the rim to prevent the motor from spinning. As torque increases so does the normal force and therefore friction between the drive wheel and rim. With two motors the forces cancel out and there is no excessive side load on the axle bearings.
I'd imagine if you were to pedal faster than the motor could spin (theoretical) there would be a torque reversal (motor braking) that would reduce the normal force/traction until it slips.

so friction drive wasn't dead enough that someone had to try and reinvent it (rubbee) but now someone went and jumped the shark.
impressive technology. ridiculously stupid application.
rims are for transmitting compression of tension spoked wheels to the ground through the tire ( though you can ride 'the rim' without a tire)
secondarily ----rims can be used as an imperfect breaking surface for applying torque.
NOW rims have ALWAYS BEEN A BAD WAY TO BREAK. hence the invention of the -----fixed cog. (technically you could break using fixed cog before freewheel was ever invented) -------foot drag break ( early 'bone shaker' design you push and break with your feet dragging hte ground. people and kids still use this method ---COASTER BREAK inside the bottom bracket pedal axle. ------DISC BREAK MOUNTED IN THE FREEWHEEL.
the problem with applying breaking torque to the rim is that the rims primary purpose is to bear the compression load transmitted through the tension of the spokes.
that is the PURPOSE of a bike rim.
when you use that rim for a frictional surface for breaking. you get a whole slew of problems with BREAKING and with THE RIM.
ordinarily the problems are solved through continuous maintaince of the breaks and of the rim (truing the spokes and hammering out the rim)
now-----------these problems aren't bad enough so this inventor now wants to put MORE torque on the rim. not to mention on the seat stays.
maybe genius, but not wise.

I love the smell of burning rubber in the morning!

Wow, zevulon. Not only is your ramble incomprehensible and completely wrong (you really should talk to somebody who actually knows about bicycling engineering and science, like David Gordon Wilson, author and emeritus professor of engineering at MIT), the word is "brake." A word to the wise, even if a word is spelled right, mixing in all caps HERE and THERE usually makes someone look like a raving lunatic.

Daishi,
Actually, electric motors are more efficient at high RPMs. Normally, you would need loss-inducing reduction gearing to get motor speed down to a good speed at the wheel, but friction drive sidesteps that. Although personally, I prefer mid-drive designs that take advantage of multi-speed bicycle transmissions.

I wonder how much lighter - and thereby efficient - would you get by using a set of 4 of these on one of those Rhodes cars (4 wheel 1-4 passenger bikes)?

Its brilliant and because its a system that assists the rider rather than being used for primary power, I see no issues with wear or efficiency. I's love to see the looks on the faces of those cocky 14lb carbon-fiber riders on the way up Lion's Gate Bridge or even on the downhill run through Stanley Park as they shoulder check only to see my fat-tired mountain bike drafting close behind...

Thank You Gadgeteer, While this approach still needs work it is not necessarily all that bad. A far more elegant design still needs to have an attainable price tag to be a success.

This might be fine on the flats, but I can't imagine it has much torque to assist going up hills.....