At the risk of stepping into the fray (I enjoy it: it keeps things lively), I think it's important in this discussion to distinguish between engineering approaches that result in measurable differences--and there are plenty of those-- and those that actually produce distortions that are audible with music programming.
BBIBH cites a number of CD-player design issues that he states "are both audible and measurable" and that "affect the sound." However, while I am not an engineer, I do have several decades of experience participating in controlled listening tests, many of which were/are conducted in one of the world's most sophisticated acoustic laboratories--the Acoustics division (Physics Dept.) of the National Research Council of Canada in Ottawa, Ontario. During these tests, and others conducted outside of that facility but in the same controlled manner--e.g., instantaneous comparisons of CD players at equalized levels playing duplicate source material, where the cost and brand of the players are concealed from the listeners-- a number of truths have emerged. This is NOT anecdotal information, or casual opinion.
Some of these so-called "problems" of CD player design are somehow compared to authentic analogue distortions such as wow and flutter, when, in fact, in the context of digital recording, data storage, and playback, they are utterly irrelevant, non-issues.
For example, vibration and rigidity are highly important in analogue gear, where mechanical and acoustical feedback from loudspeaker reproduction can be transmitted directly back to the turntable base, platter, and phono cartridge, which results in clearly audible rumble, low bass coloration, woofer "flutter", and the like.
But vibration and rigidity in a CD player (apart from the violent shaking given portable players by joggers and athletes, which knocks the laser assembly off course for sometimes seconds) is a red herring. It has never been demonstrated in a domestic player to have any influence on sound quality. Error correction is always working in any CD player--there is built-in redundancy in the Reed-Solomon code that repeats data constantly. When error-correction runs out of data to fill in the 1's or 0's, you'll hear a click or skipping. It's usually damaged CDs that cause most lapses of error-correction.
Vibration and rigidity are specious issues with CD players. Likewise, "clock jitter", while it can be measured, only produces slight increases in distortion that remain far below the level of detection with music or test signals. "Jitter" is a favorite buzzword amongst untutored audio enthusiasts, many of whom write for high-end journals. It reminds them of analog flutter in a turntable or tape deck, and analog flutter is horribly audible--yes, BBIBH, you're right, I can teach anyone to hear flutter from analogue turntables and tape transports--so therefore "jitter" must be audible in a CD player, right? Wrong!!!
But the late Peter Mitchell, an astrophysicist and audio writer for many years, pointed out the nonsense inherent in discussions of clock "jitter". And I have never, ever, heard any audible manifestations with musical programming of clock jitter. It's a non-issue.
Analog circuits and output preamps are entirely different--subject to noise, harmonic and IM distortion, just like all analogue amps and preamps.
Digital-to-analogue converters (DAC's) and ADC's, while quite variable in their design parameters, may produce measurable distortions that are audible on CD test signals, albeit at extremely high volume levels. But in my experience, and those of several of my colleagues with impressive credentials (PhDs in engineering and psychoacoustics) such distortions are always masked by any music because they occur at levels 80 dB or more below programming.
All of which relates back to Ian's comments. The differences in a transducer (a speaker, microphone or phono cartridge) and its distortions are so huge relative to the residual (and inaudible) distortion artifacts from different CD players or electronics, that it's imperative to always concentrate on the weakest link, the loudspeaker (or the phono cartridge). (By the way, lots of studio microphones are subject to the same frequency response errors of loudspeakers; sadly, we can't do anything about those aberrations other than avoid CDs recorded with "hot" mikes that emphasize midrange frequencies). The latter problem is endemic to lots of commercial pop recordings and even some classical discs.
In closing, think about this: a greater than 2-dB variation in frequency response through a speakers midrange (where most musical content resides) has a profoundly greater effect on the tonal balance of the music we hear in our living rooms than any multiple of so-called digital "distortion" attributed to DACs and solid-state electronics. Just shifting your main left and right speakers a few feet one way or the other in your room introduces far greater audible changes in the tonal balance than any number of substitutions of $10,000 Wadia CD players over a $200 Panasonic or Technics.
No, a fine loudspeaker won't conceal analog wow or flutter from a turntable, but for heaven's sake, focus on the one component capable of introducing huge changes in the sound quality of reproduced music--the loudspeaker!
Axiom Resident Expert