@Hatsuwr, the absolute TL;DR is that suction determines air speed, airflow is secondary to suction (pressure difference by the way), airwatts is like electronic circuitry (suction ≈ voltage, airflow ≈ amp current, flooring and leaks ≈ resistance), and you need both good seal and proper agitation and effective pile separation. The best cyclonic designs are far better than bags because they doesn't lose suction unless overfilled bin or reckless usage, and finally...
Dyson motors are far too small for normal ball bearings of any reasonable size, by the way.
I doubt they are too small for commonly available bearings. McMaster has 1/8" and 3 mm OD for sale, and it's not hard to get smaller from a manufacturer.
Although that does bring up a correction I should make: at this size, speed ratings of 100k-200k rpm aren't as uncommon as I first assumed.
https://www.mcmaster.com/57155K339/
Regarding your other point... Actually I'm not sure what that point is. Is it just another opinion about cyclonic separation vs bags? Not really sure why we are addressing that one again... They each have their pros and cons. Pressure loss is just one of the differences, but personally I think there are more important distinctions. Besides, pressure loss is going to depend more on the specific implementation of the system rather than just the decision about which method to use.