Here's an idea that I've been pondering for years now. I doubt if I'll ever build it myself. I just don't see much need for it for me. (though now that I'm feeling inspired to maybe some day engineer my own e-bike, who knows?)
But it ought to be cheap enough. And some of you guys who are technically oriented might be able to make some use of it.
It's a bicycle dynamometer. I'll try to attach my crude drawing of it as well.
The bicycle will need to be suspended above the ground. Right behind the rear wheel you'll want to place the equivalent of a child's see-saw. (Though it won't have to be that heavy duty. I'm sure a 2 X 4 will be more than enough) Attached to the bicycle end of this see-saw, or better yet 'the torque arm', would be a standard bicycle brake. You would slip this over the rear tire and line it up with the wheel rim so that it is ready to stop the wheel spinning. At the other end of the torque arm would be any old bathroom scale. Just have it sitting on a platform right below the end of the torque arm.
Start up the bike at WOT. Start applying the brake. (It would be very helpful to have an ammeter measuring the current running through the motor) The torque arm will, naturally, pivot on it's fulcrum and push down against the bathroom scale. You'll be able to calculate the torque at the wheel's rim, or where the rubber hits the road, from this. With an ammeter in the loop, you'll be able to measure the drive train's torque from 'almost no load' right up through 'watch it! you're gonna burn up the motor!' With a tachometer (or a bicycle speedometer) you could measure power through this range as well.
For anything more than quick tests, heat will be a problem. Might melt the brake pads. But if you ran water over both sides of the wheel rim right before the brake pads, then you'd have all the cooling you'll need. I suppose it'll also make the braking, and measuring, more smooth.
In my drawing I put the fulcrum at the wrong end, now that I think about it. Imagine you were grabbing onto a wheel that was delivering, say, 50 ft/lb of torque. If the fulcrum were 5 ft from the wheel and 1 foot from the scale, then the scale would be absorbing a 'weight' of 250 lbs. This would put your measurements right in the useful range of that scale. I'm sure that you would want the fulcrum to be closer to the scale than to the bicycle.
I suppose this could be useful to someone who wants to engineer the most efficient drive train.