Dimples only work on subsonic air flow. Once the speed of sound is reached, the mach waves (a cone for a sphere) attach to the object and the aerodynamics change significantly. The simple Bernoulli equation of flow fails as soon as standing pressure waves occur.
The device will produce gyroscopic force, the question is how much.
There will be ‘recoil’ of one form or another since momentum must be conserved. The recoil should appear as a loss in angular momentum of the spinning disk that imparts energy to the projectile. If the loss of momentum is made up by the motor driving the spinning mass the recoil will appear here. You can move the force around, bt you cannot eliminate it.
The device will have a heat signature since it is using a motor and dissipating power, and it is going to need a lot of power.
Very high speed ball bearings typically have a relatively short life without a good supply of pressurized cooled lubricant. Magnetic bearings might be a possibility, but unless they are based on permanent magnets take a significant amount of power to operate.
There is going to be either a lot of noise generated when the projectiles are quickly accelerated or a huge gyroscopic mass if the entire supply is accelerated and then metered.
The energy from a 150 gr projectile at 8000 ft/sec is 21,313 ft-lbf (grains are not a unit if mass, slugs are the unit of mass).
This is about 29 kJoule per projectile, and at full speed 120,000 RPM, 2000 shots/sec wil take 58 MW. And this is before any losses.
This will not be a very portable item. The power will scale with firing rate, but even 5000 RPM will require ~2.4 MW. A rather large power requirement. To place it in perspective, this is around 3,240 HP, so maybe a very large turbo-generator. And a very large vehicle.
Maybe the platform should be the world’s first nuclear powered tracked vehicle. We did it with ships, why not land vehicles?