Here's what actually happened....
A couple of the track names on the Sooloos were in an old format and they looked old to me. I was worried they might be very early (2013) experimental files. I said something like, "Are these the latest?" They were in fact all 0.9 direct from the Shard demo i.e. the latest.
The files on my laptop were from a copy of the 0.9 files used for headphone demos at the Shard and were also in use on the What Hi-Fi booth at Bristol.
The 0.9 audio standard actually pre-dated the Shard by many months. Only those involved in the development of MQA have heard any other versions, and as Bob points out [in a private email], later versions will sound the same.
I had intended to play some additional 0.9 files from the laptop that are not on the Sooloos, but in fact that did not happen because there were too many questions.
Yes, there are some differences in what this 818 displayed when the USB and Sooloos inputs were used. It's a prototype system.... Don't believe the display unless you know what the software behind it is doing.
I did comment that a momentary display of the sample rate on the 818 of "48kHz" indicated both the low data-rate of encoded MQA and the possibility of backward compatibility to 48kHz capable systems. There's nothing new in that. I said it in the slides and Bob said it at the Shard.
Of course everything I actually played was decoded full-quality MQA as indicated in plain view on the DSP8000s.
Feel free to quote the above.