Your statement is undeniably correct.
At the risk of contributing to making this thread grow like the ones on the Roon forum, I would like to expand a bit on what I meant (non native speaker here).
Quote from the article:
Taking all of these things into account, the MQA format approaches the efficiency problem from the standpoint of matching the encoding format to the capabilities and requirements of the human auditory system, focusing on the relevant and wanted information without wasting resources on the irrelevant aspects.
The relevant words been “irrelevant aspects”.
Why is this discussion of lossy vs. lossless a subject of such controversy. It’s because the two camps use “lossy” or “lossless” in different ways.
When I say lossy, I mean that the complete high resolution PCM source signal cannot be replicated from the MQA file. Contrary to FLAC compression / decompression, the MQA origami cannot be reversed without loss of information.
When Bob Stuart says lossless, he refers to his opinion that the unfolded stream is perceptually indistinguishable from the high resolution PCM file, hence in his definition “lossless”.
No wonder this discussion can never end, as both sides implicitly use different definitions of the word lossless. MQA are very careful to never making their definition of lossless explicit.
There is a second meaning of lossless that MQA use and that is perceptual losslessness (=accuracy) over the whole recording / reproduction chain. See Bob’s discussion of the white glove treatments, where certain errors of the recording process are corrected during reproduction.
I find this white glove approach truly fascinating. Unfortunately it can only be used in very few instances. Actually only two examples were ever done.