You misunderstand the physics involved.
It's a physical property of the motor. When the controller applies a voltage to the terminals of the Permanent Magnet DC motor, current starts to flow in. As the motor spools up speed, it automatically starts to generate an opposite voltage to what the controller is pushing out. The back-EMF voltage. These two voltages "battle" each other so that when the motor gets enough speed, the voltage it generates equals and opposes the voltage of the controller and thus negates it. No more current flows in, no more power is generated and the speed cannot continue to rise.
This is the situation when the car has reached its travelling speed after acceleration. Now of course, the air resistance and all that drags the car and the motor back, so the speed of the motor will never rise high enough to completely nullify the voltage from the controller, which is why the motor continues to draw current and consume energy to move the car. The system goes into a balance where, if more load is put to the motor, it will slow down and generate less backwards voltage, which results in more current flowing in and more torque being generated to match the load. It means that the back-EMF is never higher than the battery voltage. For that to happen, you would have to travel faster than the highest speed you can accelerate to with the battery voltage.
Now, if the controller should suddenly stop putting voltage to the motor, the motor's own backwards voltage would still exist. The controller has to shunt that voltage somewhere to draw out the current. If we weren't interested in taking that energy back, we would simply direct the voltage around and short circuit the motor, which means that the energy would be lost as heat in the resistance of the coils. The motor would start to brake rather hard in that situation. Or, we could simply ignore the voltage, which means that we open all the switches and let no current flow anywhere. The motor continues to wheel freely and a voltage exists across its terminals as if it were a battery itself.
If we want to take that energy back, we have to direct the voltage to somewhere we can make use of it. We can't put it straight into the battery in any case, because the battery voltage is higher and that would make the motor turn even faster when we're trying to slow it down. We have to direct it somewhere where the voltage is lower. In theory, we could put it into a 6 volt battery which would then charge up happily until the motor turns so slow that it generates precisely 6 volts where the charging stops. This approach has its problems, because the huge voltage difference at start would probably cause the battery to explode in a sudden rush of high current, and the passengers would be slammed against their seatbelts due to the rapid deacceleration. The same problem would occur should you place a capacitor in there. The voltage of an empty capacitor is zero, so while the capacitor would survive the huge inrush, the passengers would have their noses against the windshield.
So in practice, there is an inductor. A simple coil of copper wire, perhaps with some iron in the middle. An inductor resists the change in a current going through it so when you connect it to the motor, it takes some time for the current to build up. Conversely, when you disconnect the inductor, it wants to maintain that current it already has and generates a huge voltage spike up to thousands of volts in an effort to find some route for the current to flow. This is because the inductor stores energy into a magnetic field when the current builds up (but not when the current is steady) and when the current stops, the magnetic field collapses back, releasing all the energy in the form of an electric jolt through the coil. This is how the basic boost DC-DC converter works.
Now, all you need is a capacitor that can take in such a huge inrush. And where you find such a capacitor is in the motor controller's input filtering capacitor bank that buffers the power from the batteries for the motor controller so that the batteries don't have to suffer from the jittery and choppy power draw of the controller, which is why they're excellent for filtering the choppy and spiky power surges from the regen brakes, smoothing down the thousand volt surges into something more manageable. As those surges come in one after another, eventually the capacitor bank voltage rises above the battery voltage and current simply flows back into the batteries in a steady stream until the capacitor voltage goes back down again. Now that you have the inductor and the capacitor, you can keep flipping the switch, connecting the inductor to the motor, then the capacitor, then the motor and so on, and it will pump the energy out of the motor and back into the electrical system at a controlled rate. The practical limitation to the pumping is that as the voltage from the motor decreases, the rate of energy you can pump out decreases while your switching losses do not, which is what I explained previously.
So you see, it is not my intention. It is how the system already works. In practice, the controller uses both inductors and capacitors to control the rate of current moving in the system so the whole thing doesn't just jam in a split second and burn everything up. There's very little to be done by adding supercapacitors in the mix, because they have properties that aren't suitable for the task in hand. For example, the equivalent series resistance (ESR) of a supercapacitor is somewhat higher than of a conventional capacitor, which makes it less suitable for power filtering tasks. That's because they resemble more a battery than a capacitor. And modern day batteries are more than adequate to handle the braking power in a car. If they can supply half a megawatt in an instant, they can take half a megawatt in an instant.
