Sure The BNC to BNC connection is RF sensitive; and it's due to the RF noise that the M scaler creates, rather than Dave being sensitive. USB is insensitive to RF, with the listening I have done on my sources. The only time I have heard changes with USB is with audiophile cables that actually create more RF noise, and sound worse. But certified class 2 cables sound the same.
The M scaler, with it's huge amount of processing, does create ground plane noise, which is subjectively worse at 2 GHz or thereabouts. The nature of this noise makes it particularly a problem, as it is weakly signal correlated; and this can interfere with depth perception. Now the BNC OP is galvanically isolated (the grounds are not connected - except by 5pF of capacitance). Unfortunately, it's this 5pF capacitance that couples Dave ground plane to Blu 2 ground plane, so this 2GHz noise can become a current that flows through Dave's ground plane, and back into the mains. By using ferrites, that are at 2 GHz or designed for GHz rejection, increases the impedance, and so the currents are reduced. This is why simple clip-on ferrites ferrites work with Blu 2.
But with the Hugo M scaler, I improved the isolation - in simple terms, put 2 GHz chip ferrites within the BNC driver (plus some other things that are too complex to explain). Now with the M scaler, you will actually degrade the SQ if you use clip on ferrites - but the solid core ferrites do offer a small advantage.
Using longer cables works with both Hugo M scaler and Blu 2; again what is happening is the extra inductance of the ground return increases the impedance and reduces the size of the current in Dave's ground plane. Going longer than 2m does not give any significant benefit - also the cable should not be looped...