Bonding should be by the widest practical means. Wide cable tray is effective when it is made of zinc plated steel and carefully bonded at the ends to control panel and motor frame. Zinc plated sheet steel channel is also effective. The fact that the width is folded into a U shape does not matter. A closing lid helps. Solid steel conduit bonded at both ends is effective. The spiral construction of flexible conduit makes it less attractive for RF shielding because the spiral shape forms an inductor, even with partially shorted turns.
The PWM Drive (pulse-width modulated drive) to motor power conductors are typically the most intense noise source in a system. Proper implementation of shielding, grounding, splicing, and treatment of excess cable is essential to reducing noise in your system.
If the coil of wire could carry on moving like this, it would rotate continuously—and we did be well on the way to making an electric motor. But that ca not happen with our present setup: the wires will quickly tangle up. Not only that, but if the coil could rotate far enough, something else would happen. Once the coil reached the vertical position, it would flip over, so the electric current would be flowing through it the opposite way. Now the forces on each side of the coil would reverse. Instead of rotating continuously in the same direction, it would move back in the direction it had just come! Imagine an electric train with a motor like this: it would keep shuffling back and forward on the spot without ever actually going anywhere.
As a technician, I naturally look at things from a nuts & bolts perspective. As technicians, we understand objects and assemblies. We understand how things come apart and go back together. Is not this what Einstein and Ohm did? They figured out how "it" comes apart, which led to the ability of others to put "it" back together.