The man relying on the "I'm a Musician™, so I don't have to understand the science" and the resident contrarian population jumping in to peddle their usual "You see? Flat magnitude and smooth directivity are not objectively good!" is something else.
I'm not sure who exactly you're addressing, but I'll volunteer to answer. I'll start by saying this is the kind of discussion that needs mutual goodwill and tolerance as we seek an answer to a complex question, rather than the kind of antagonism and aggression you include above.
I'm a musician, and I also understand the science pretty well. Probably better than you understand music, anyway. I'll start with an imperfect comparison: low noise and distortion are important in electronics, but improvements that happen below the threshold of audibility don't matter; just as FR in a speaker doesn't matter beyond a certain point.
The former is well understood and widely accepted. The latter hasn't been tested yet. Except it has, billions of times over the last century. Musical joy and ecstasy have been achieved by listeners with grossly inadequate equipment, when compared to Genelec standards. How is that possible?
We know the brain has immense audio processing capabilities. We experience the effects every second of every day. The sophistication is stunning. Is it unreasonable to surmise the brain interprets and smooths FR where it needs to? In a way we can prove it does - think of the serious record reviewers from the past, when that was an honorable profession (still is, mostly) - how could they make musical sense of what they were hearing, if we judge solely on technical grounds? How did they sense touch and nuance, and line and flow, when we know the sound power in their room was hopping randomly up and down?
My experience over 44 years as a pro on both sides of the glass is that all speakers are pretty inadequate. All of them eventually reveal their clumsy, lumpen, mechanical nature. At best I carry around a mental database, a bit of this, a bit of that, a kind of imaginary Franken-speaker that would combine the best parts of the best I've heard. Some of those parts are cabinet silence, headroom, lack of compression, low distortion, bandwidth ... undoubtedly all measurable metrics, yet they're rarely reported and it's rare that they're considered as a related suite.
But if all that's not science-y enough for you, concentrate on explaining to me how one somewhat-wiggly FR is better than another somewhat-wiggly FR. Is it where the wiggles are, or their height, or spacing, or exact shape, or what? If you can't tell me, will you admit there
might be more to speakers than spinoramas?