My tests got completely disrupted when I moved my hifi rack to a different room (mostly for WAF kind of reasons).
The key difference was that the orginal location had a dedicated mains line with high quality off-the-reel fat Furutech cable - and the new location had just the regular mains ring that goes round the whole floor of the house. But in both cases, I had a high quality power line conditioner (Audience TSS), so I reckoned the move wouldn't make much difference.
But I was wrong. one aspect of SQ seemed to go back 2 years of careful progression. Detail and dynamics were fine, but the problem was the re-introduction of some mid/HF glare and brightness from my laptop that I had previously managed to reduce. I tried various tweaks to see if I could find the problem, which maybe changed things subtly, but the change that made the big difference was swapping the mains cable to the ISO Regen:
In the old location, the IR was powered by MCRU connected to my Audience PLC via a (2 generations old) Transparent cord. In the new location I used an Audience Powerchord e (that I had considered to be slightly better than the Transparent based on a test in a different situation long ago). When I swapped the Transparent cord back to the MCRU, the glare seemed to be tamed again. Phew! Because of the above setback, I had now lost my previous reference point, so I can't reliably say if the sound was exactly as good as in the old location, but at least it was in the same ballpark.
So, with getting back to approximately a good place again, and showing how important power supplies are to this situation, I tried another swap: Give the MCRU to my mR and give the higher quality LPS-1 to my IR. Result was mixed: Possibly the best clarity/detail yet, but also leaner and with some of that glare re-appearing. So I'll swap this back again and then try taking the mR and network out completely, using USB direct with the IR. Sadly, there's still no sign of the Paul Hynes SR7 - now over 2 months later than Paul's ETA estimate and he didn't reply to my last email. Grrrr!
The moral of this story is that seemingly irrelevent changes can skew any result. I'm not surprised that there are so many different reactions to the IR or any other component.