Agreed on points 1 and 2.
As to RF filters, this is a complex area; also 22dB reduction would not be enough to eliminate the issue; also it's rated using a 50 ohm load which is not what one would get in practice. That said, best thing to do is to give it a listen; if it sounds warmer then run with it. I had no benefit with using batteries on TT2 though, but each system is different.
As to BNC cables, clip on ferrites helped with Blu 2; but they sounded worse with the Hugo M Scaler, due to the improved isolation on the BNC outputs. A simple way to improve the SQ is using 2m of BNC cables rather than 1m. The cable length is important (longer length is greater inductance to 2GHz impedance for ground loop currents is greater), so to a much lesser amount cable construction. Solid core 2GHz ferrites do work and will give the largest improvement and I use Wave cables myself.
Running the M scaler on batteries does indeed improve the situation and solves the BNC cable issues as batteries (you need to run it isolated completely from the mains) severs the ground loop from M scaler to TT2; thus RF currents can't flow into the DAC ground plane, and the problem is solved.
I should stress that this is very much surface finish of the icing on the cake - it's a significant difference on an AB test, but the scale of the change is forgotten about after a few days.
Happy listening!