Look for something on this subject from George Rankin (wavelength), is where I would start.
I expect we are looking at a re-clocking issue of some sort.
The vast majority of older CDP's, IIRC, get some of their important aspects of clocking cues from directly off the disc read. I has to do with the original design spec. Ie that the disc read itself is intrusive into the jitter spectrum question. Part of why Sam Tellig of Stereophile thought that some of the first buffering portable CD players sounded quite good. The buffer means the disc read jitter spectrum issue is not directly connected to the emergent waveform of the music signal. That sort of hardware does a re-clocking of a sort.
When we rip, the clocking happens elsewhere and if done well, it can be a better reconstruction of the waveform, jitter spectrum wise.
The given individual case sensitive final clocking set up and reconstruction of the waveform is where it is all at. Which is pretty well what Gordon Rankin is on about.
Note to Geoff: the Popper reference elsewhere made me look him up. Thanks for that. I (and many others, obviously) have similar or parallel conclusions in many ways, but not so explicitly perfected and outlined. I've not spent a lifetime refining it as he had. It's nice to be confirmed in one's ideals/conclusions.(Standford website, I'm still reading...)
And the more I say the more I become a target, which takes you 'round to his thoughts on theories......