I think these ideas are right. If the volume on the ZMA is wide open, it is not being adjusted by the pot, so is like no pot volume-wise. Then by losing the wires and pot, there is less in the signal path, so cleaner.
It occurs to me that one consideration here is voicing. If Steve designs it with the wires and pots for that convenience, with minimal signal loss by short path design, good materials, etc...And since these elements are part of the "voicing" of the amp...By removing them, the sound will be more "purist." But if you loved the sound as designed, by removing, or "improving" some parts to make it cleaner, there would be the possibility that the sound change could feel a little too clean?
For example, I have the CSP3 setup with my Torii as one, compensating for the longer signal path with specifically chosen "cleaner" tubes, cables, fuses, feet, etc to make the whole sound "transparent." It is not strictly as transparent as it would be with no CSP3, but due to the cleaner compensations, it would likely be too clean as setup now with the Torii only. And I would lose the OTL qualities of the CSP3 and gain riding....adjustment I do to alter signal density based upon recording quality....one way, leaner, and the other denser, easily fine tuning for taste for each recording. And gain riding for the most part requires a volume on the amp and pre.
So if the ZMA sounds "right" to Steve as designed, "purifying" some of the signal path might skew the amp some toward cleaner than Steve might have chosen for production. May not be better or worse, but different, and then the taste thing.
If a DAC has a pre, it can be in effect like a separate pre. Then it depends on the pre circuit quality and sound as to whether or not it might sound best or not...including better or worse than the ZMA volume pot circuit.
Technically, the ZMA with no pre would be a cleaner signal because there is less for the signal to travel through. Any wires, resistors, caps, etc that the signal travels through effect the signal...but then the question becomes: what is the end result of a signal modifier? And how transparently does it modify the signal?
In the case of the Zstage, CSP3....or Zbit, (volume adjustment abilities of the first two aside), how they are built and how we set them, changes the signal on purpose. There is inherently some loss due to a more complex signal path from parts in the units and added ICs....and some gain due to running the signal through the tubes or transformers designed to "improve" sound quality, part of which being, making the output voltage adjustable....allowing our adjustment of signal density.
At the same time, many with this level of gear purposefully "calm down" their systems a bit here and there with ICs, tube, feet....choices with sound signatures that are not the purest, but create the most pleasing sound to the individual in an individual system and room. And some purists have limitations created by less-than pure system/room anomalies (known or unknown) that cause them to need to stay as clean as possible. Likely room for most, but also using ICs, transports, power cables, DACs, feet, caps, resistors, speaker cables, etc that may not allow the purity of their amps or front ends, or whatever to be fully revealed.
Especially any part of the front end can limit the potential of a whole system. But also, one set of ICs can be limiting from the IC onward, effecting the "purity" of the whole....one tube.....the purity of power and power treatment feeding gear...power cords....vibration or low quality vibration "adjustments"....lots of ways to never get all the purity potential that is already available from the best aspects of our systems and rooms without mods.
That said, I personally am pretty deeply into modding to taste, purity and completeness of natural sound qualities primary players in this...