USB audio is defined by USB Audio Class 1 (1998) and UAC2 (2009)
The first uses USB in full speed (USB1) hence it is limited to 2 channel 24 bit 96 kHz sample rate.
The second requires high speed (USB2) hence no limitations on sample rate.
It also took some time before Apple and Linux included UAC2 in their OS (2010) not to mention Microsoft (2017).
This explains why on older models USB is capped at 96 kHz.
Bit more detail: https://www.thewelltemperedcomputer.com/KB/USB.html
One thing to do when comparing highres to CD is making sure you are listening to the same master.
Your best bet is to use a highres you own and down-sample it yourself to 16 bits / 44.1 kHz.
Before you do make sure the highres recording indeed contains information below -96 dBFS and above 21 kHz. Not all of them are genuine.https://www.thewelltemperedcomputer.com/SW/AudioTools/Spectrum.htm
“ tout improved dynamic range which is the counter to compression I believe”
Often people mix up dynamic compression (decreasing the difference between the loudest and the softest passage) with lossy compression, removing information to reduce file size (MP3 style).
Remasters, often in 24/96, often have less dynamic range then die original recording as released on CD. This is known as the loudness war. This yields the paradox that in these cases the highres version often sounds worse than the original.
Oh yes and volume level does make a difference.
Our hearing is very un-linear and to make matters worse, this varies with the loudness.
Known as the Fletcher–Munson curve.
This is why matching volume is so important when comparinghttps://www.thewelltemperedcomputer.com/KB/Hearing.htm