we ran our own 'matrix test' as we called it over here a few years ago. We wanted to se if we could duplicate those results.
Whilst the figures are not exactly the same (duhh) we did in a general sense duplicate the flavour.
We went as cheap and nasty as we could on the 'poor' front end, very much 'upgraded' on the superior. I used house wall wiring for speaker leads on the cheap for example.
One additional thing we did that matrix did not was that for the part of the demo we did (we were supposed to do 'round two'..but due to a fvck up on my part grrrrrr, am so annoyed by tha still grrrr....we did not manage to do it) we had the analog output from the cheap end going into the adc of a behr 2496 (bypassed) and straight back out the dacx then to the speaker.
In other words, unlike the matrix test an additional adc-dac step in the cheap end. (the second part was to engage the 2496 and have room eq etc etc)
MOST who expressed a preference did choose the 'better' one, so in that sense not the same result. Those who want to belive can take comfort from that!
The truth is that all were shocked at how damned close they sounded..and I mean damned close. AFAICT the 'quickest' anyone could even begin to distinguish the two was about twenty minutes of concentration. (unlike the matrix test, each was alone in the LP for as long as they wanted, and could switch at will between the two).
Only one person got to hear the expensive front vs the cheap front with the 2496 activated (ie befoire the fvck up) and he instantly poreferred the 2496 side. He previously had walked away from the first part saying 'they sound the same to me'. I watched him do that, he listened the first time to five or so tracks and could not tell them apart, have the room fixed etc with the behr and he heard and preferred that after ten seconds.
I am still kicking kyself for mucking that second part up. Anyway, all agreed that there was absolutely NO 'night and day' difference. It made more than a few audiophiles very uncomfortable.