Once again, I really appreciate you putting together that comparison, and especially how much work it took. But I think we need to be very clear that the comparison you set up is, with respect, meaningless for determining speaker preferences, for the reasons noted by several folks in that other thread. I don't mean this to be unpleasant or disagreeable - again, I appreciate the comparison and it was a lot of fun! But to say there are a number of additional, confounding variables in your comparison is to make a massive understatement.
To be clear, I invest in speakers with the best preference score too. It’s the same rationale as electronics: since it doesn’t cost a lot to get SOTA, why not?
But when you look at this
There are things that we should consider:
1) We know that bass plays a big role in preference
2) The original preference score correlation was done with smoothened FR measurements and may not be as predictive with increasingly precise measurement methods
3) We have adopted the philosophy of mono evaluation is the best way to evaluate speakers. On the other hand, an equally scientific conclusion is: “In stereo, tonality issues are masked. Thus there must be a point where the preference score is “good enough” and it would be preferable to invest in better bass with worse tonality than smoother tonality with less bass.
4) We know a horrible speaker in a great room is bad.
But is a good speaker in a great room or a great speaker in a good room preferred? Is there a crossing point between good and great in either axis?
These are questions we should be asking. My blind test not being sufficient to address the answers to these questions, but hopefully demonstrating that these are simply questions to consider.