I spent another couple of hours searching last night and found nothing. The fact that nothing like it exists on some very complete parts vendor's sites for all things minibike and gokart makes me wonder if, like you suggested, it might be too complicated to build to compete with the centrifugal clutch (at least for minibikes and gokarts). Centrifugal clutches can be pretty darn sophisticated, with a range of rpm curves and lockup rates, they fit the bill nicely enough for even the most technical racers. For these single ratio transmissions a manual clutch probably wouldn't offer the rider/driver any real advantage since the power curves of the motors/transmissions are pretty calculable and constant.
But for motorized bicycles it would be fantastic. I'm looking at that shift kit from sickbikeparts and dreaming of putting a little manual clutch if one were available as I described on the right hand side of that jackshaft. As such, you could dispense entirely with the centrifugal clutch at the motor, and as a bonus, the entire motor drivetrain would be isolated while pedaling with the motor off.
I don't know.. two sections of a clutch locked separately to a shaft with an idling sproket between them. One of the plates is fixed via arm to motor or mount, and within is a round pressure plate that is driven into the idler (both having thrust bearings) via coarse threading or multiple ramps around the plate. When a plate is rotated with the cable, it is pushed away from the other plate into the idling sproket/clutch face. Easy enough to imagine. Kind of suprising that it doesn't exist in some fashion out there.
Heck, the blade disengagement clutch under my honda push mower is extremely close to what I've described, but about 2/3 overscale and overweight to put on a little trimmer motor. It engages and disengages with a manual cable (unlike 99percent of the riding mower blade clutches out there that are 12vdc). So I know the concept is valid. I have a hunch that it exists on some product out there. I just can't picture the application. Economies of scale can be a real boon to experimenters, if you know where to look. When something already exists as an automotive part or home durable goods part it's going to cost 1/100th of the price of something needing to be custom built. Often, it's just a matter of figuring out where that part is already in use.