Seems like a parlour trick to me designed to confuse people and mask real world and even audibly measurable differences, let’s see how we can get NULL results on everything possible+1
Audio is not life and death, so there is nothing seriously wrong with following subjective impressions, but only as long as you remember that - that is all they are - subjective impressions.
As are the impressions of those under test. Their hearing is still completely subjective whilst the testing is being touted as objective. There's a lot of conflating going on there. The parlor trick of A/B testing can reduce any desired result to that of chance. You're not listening for pleasure. You're listening under test conditions. No matter how "objective" you think you can be, that very aspect is constantly weighing in on your conclusions.
The very nature of a test implies the ability to discern so it's inherent that you expect change. Now, factor in the "cheat" of implying a change, only to not do so. That is the reason the mind may perceive a change in a quick test scenario. It's just a trick.
Any legitimate evaluation is done over the long term with equipment and music you are familiar with. It is then that you can be comprehensive and confident in your evaluation when you change something.
All the best,
Nonoise