If you put 44.1KHz music into Davina and set it to output 44.1KHz music, is it possible that it could be "better"? I'm thinking that Davina would do 1M-tap upsampling to 705.6KHz then dither/decimate to 44.1KHz and in the process produce a file with "superior" encoding at 44.1KHz due to improved dithering.
This might have to be a two-step process, producing a 705.6KHz intermediate file at step one, to be dithered/decimated to 44.1KHz in step two.
In theory it would be simpler to use Davina to decimate to 24-bit and that would sound better. Dither is simply a "partial fix" for the resolution lost in 16-bit encoding (and decimation, if there is any), versus the "24-bit" input.
There is, unfortunately, one more element in this chain of logic: the difference may not arise with insufficient ferrites (or if they're badly configured - they should all be as close to DAVE as possible), because there's a threshold effect. And because the resonant frequency of the ferrite(s) alters how much filtering occurs in your particular system.
So, while it's possible to hear an improvement with some ferrites on the cabling, some people might not hear a difference because their existing cable is actually good. In that case they would need to use more ferrites, and they might run out of cable length before hearing a difference.
Now playing: Ani DiFranco - Amazing Grace