I have to gently remind you that you only contacted 2L after months of cojoling from me, and over 5 years since the files were made publically available by 2L and MQA.
So welcome to the world outside the "Flat Earth MQA Society," as it were.
Perhaps in return I can remind you that I tend to do things in my own ways. And a part of that is intitally to investigate the actual items of interest as the basis. I had in mind to contact people when/as it might seem useful. But do one thing at a time. I 'work' for myself, not you or 2L or anyone else here when I decide to investigate something that interests me.
My main intial concerns were:
1) That the effect of MQA encoding on Audio CD might be large enough to significantly degrade *non-MQA replay*. That would be a concern *regardless* of the 'provenance*, etc, of MQA.
2) To explore and understand how it works in practice and cut through the claims made by its inventors, etc, so I could explain it more clearly. Also to assess if there was any real 'need' for it given free open alternative that would save people the costs and bother.
I used the 2L examples as they are very convenient, but I used them 'as given' for the above purposes. I'm afraid it is your invention that this made me a member of your fanciful "Society". I'm just doing what researchers do.
That took a few months.
Having concluded that MQA is "mostly harmless" and probably also "mostly pointless" I'm content to let others make up their own minds, taking into account what I wrote as a part of that. The issue of 'provenance' arose because of the possibility that any 'change' or 'effect' shown up by the 2L examples might be for *non-MQA" reasons.
Given that the effects seem trivially small, I doubt that matters much. But having done the above it made sense to ask about it now we had some info about the actual differences.
So I asked, to check. I have a reply, and that is useful. I see no reason to doubt what I have been told. BUT as usual as an (ex-) academic scientist I have to keep "Take no-one's word for it" in mind as people can be mistaken. But it means that so far as the 2L files are concerned we can now be confident that changes that cause a version to differ from a simple standard downsample of the DXD *are* due to MQA. Thus we can 'assign' any changes that have an impact on MQA and non-MQA playback to MQA and not some other cause.
I am now working on other things. I may well return to this at some point. But if anyone else wants they can try pre-empting what I have in mind. So I can summarise what I'd plan to do next if they want to have a go.
BTW the only 'Societies' I belong to these days are the IEEE, AES, and IoP.