First of all, to those not interested or keen on the DAC chip itself do matter in some cases, please ignore my long post below.
There is something wrong with their logic though, if that's true, because
1) We all know labor (in this case research/design/development) is a big part in the manufacturing of devices such as AVPs. So if the manufacturing is willing to spend say twice as much on the "implementation" side of the DAC, and that would get "full capability" of a TOTL , that would be a strategy welcome by those who know enough to prefer DACs that are well implemented for the better results, so why limit the potential by a DAC chip that has a spec of only 93 dB SINAD?
2) If, based on your "implemented to its full capability..., than the TOTL chip used would not be performing at 95 dB, it would be at least 110 dB (usually higher) that TOTL chip could perform. Example: the DAC chip used in the AV10/20 is the reference class, but not TOTL class of ESS Sabre and it has SINAD spec of 120 dB. That's a lot better than 25 dB.
3) Conversely, if the manufacturer picked a chip that has SINAD spec as low as 93 dB such as the one used in the midrange Denon and Marantz AVRs/AVPs months after the AKM factory fire, than in theory the best those units could perform on the bench would be 93 dB. Since the Cinema 40 and the AVR-X4800H achieved as high as 92 dB using HDMI input, we can be reasonably sure that D+M's team did a decent job in their implementation, just too bad the DAC chip in this case has become the bottleneck. This seems to show in this case the DAC chip would be the source of distortion/noise, simply because the only change in the pre out signal path is the DAC chip, based on available information published by D+M.
(1) I would think you are right about that, but your key word "typically" is key here. In this case, again, the TI PCM5102A's SINAD spec is 93 dB, whereas the other parts up and downstream, such as the Volume control chip and the opamp chips have SINAD of over 100 dB, more like 105 to 110 range. So in this case, you would probably agree that using an improved DAC chip would not be a wasted effort.
This is why I mentioned before, the "implementation" being more important blablabla often seen on forum posts, sometimes by professional reviewers are often overrated. I highlighted often, because obviously it is more or less a case by case matter. For example, I would say in the case of Anthem AVM70 vs AVM90, implementation would be more important than choosing between the ES9038Q2M (the 70) and the ES9038Pro (the 90). In the case of the AV20, likely the AV30 if they use the same one in the C30, then again implementation should be the focus (my opinion only), whereas for the midrange AVRs, the likes of the popular AVR-X3800H, Cinema 50 through 70, focussing on the chip would have been welcome by many, like the so called SINAD chasers lol..
(2) One last thing, for others who are keen on getting the better DAC chips, I would like to point out that since Amir discovered that even using the PCM5102A (SINAD spec: 93 dB), some D+M AVRs managed to achieve SINAD as high as 97 dB, tells us D+M's team knows how to implement DACs to get the full or near full potential of the chip, and/or the 93 dB spec of the TI chip is probably a typical number, and it might be (now this is just my educated guess/speculation...) possible that D+M, as a large volume buyer, managed to get TI to sell them selected chips by something like sorted bin, so they were able to secure the same 5102A, but those sorted by having SINAD x dB better than the specified 93 dB Other than this last point though, my other points are based on logic, not facts and feelings, but strictly logic.
.
