I agree with everything Rammis has written, 100%.
I knew Paul Klipsch personally. When I was a boy he and his wife would visit our home for dinner and would stay all evening talking and playing music - his wife played piano, as did my mom. My mom was one of the judges for the Klipsch piano scholarship at NMSU and my dad worked with the NMSU EE department (Paul's alma mater) to improve their work in producing qualified engineers for the work my dad did for the government during the cold war. It was a special time in my life, and very similar to the images of the 1950s and 1960s often portrayed humorously in movies and that artificially perfect world.
Anyway, his favorite listening test was to place a La Scala or Heresy under or near a grand piano and have his wife play or have a reel-to-reel recording of her playing reproduced over the speaker. He challenged anyone to hear the difference and he claimed no one ever could tell a difference. He always wanted to make the speakers now known as the Klipsch Heritage line sound perfect, and for him that was the ultimate test. But he also bemoaned the marketing department who influenced the crossover designs to reflect the trend at the moment for what a speaker should sound like to people spending that much on speakers. The company followed the "audiophile" sound through time and the hardcore fans all know which model of crossover sounds like each era. All of the crossovers are still available from Klipsch, for owners to get what they expect from their speakers.
While I was at Dell I worked closely with one of the top engineers at Bose when the Acoustimas system was introduced. When I knew him he was a CTO at Dell, but he loved to tell stories about the massive amount of research into listening preferences of home audio nuts and they developed a frequency response which fairly consistently created a sound that made nearly all recordings sound pleasing to listeners. It didn't need high output above 14kHz or deep bass below 80Hz. It had a dip in the midrange from about 2kHz to about 6kHz and a boost in the upper bass in the 120Hz to 250Hz range which, when all done properly, made just about any recording pleasant to listen to. Personally, I found it made everything sound the same, lacked detail (which has little to do with the response and more to do with the single cheap-ass 3" full-range driver and shitty semi-sub-woofers and extreme phase response problems below 300Hz). Their target response was like listening to the classic "Flying-V" graphic EQ curve, but with an extremely low cost speaker system. I still swear that everyone I know who ever purchased an Acoustimas system got bored with it after a few months and soon stopped listening to it at all, except on special "movie nights" and cocktail parties.
My experience with "Audiophile" speakers from the 1990s and 2000s is wrought with good attempts at great sound often killed with stupid crossover voicing decisions. A smart engineer with no ear for music would design a speaker which solved an often under-looked issue in performance, such as time alignment, baffle step reinforcement, or baffle edge sourced echos, but then made stupid decisions with crossovers or driver selection.
I once demoed a famous speaker which claimed to have a natural sound because the frequency response was measured to be flat in the listening space. Well, the response was actually flat, but the overall levels of the woofer, midrange and tweeter were relative level, but the crossover areas were insanely bad from poor design and the tweeter had a natural resonance at 15kHz from the metal dome breaking up. The speaker sounded very harsh and had serious issues, but the marketing suggested that's what good speakers sound like. It was sad.
Audiophile speakers with that infamous British sound had a clear dip in the 3.5kHz range as prescribed by the BBC to address a very real issue with using speakers above a large mixing console and introducing a reflection off the console often causing a boost in the response around 3.5kHz. Well, many people liked the sound of the studio monitors designed for that applications with the dip in the response and assumed that meant more accurate sound. Audiophile speaker designers would exaggerate the dip, sometimes to the tune of wide dip centered at 3.5kHz which was more than 6dB below the average response. This reduced the levels of the most important content to human hearing and identification of sounds so the brain was forced to fill in the missing information by interpolating from the harmonics above that range, leading to a perception that the listener was "hearing" detail they never heard before. When the main way to know an acoustic guitar is a guitar is based solely on the harmonics above 6kHz, that can be very sexy initially, but it leads to mental exhaustion and ultimately listening fatigue. Those speakers also tended to have a strong boost in the upper treble, somewhere between 14khz and 20kHz.
In the 1980s Dr. Floyd Toole introduced the well-researched "Canadian" sound with the Energy-8 concept speaker which later was used as the starting point of the Energy speaker company and similar sounding speakers using that research started up, such as Paradigm and others. That sound quality was based on a smooth response, consistent propagation (dispersion) and an in-room response which sloped down from about 600Hz to 20,000Hz. For a period that sound was taking the world by storm, but competition in the store taught everyone that in listening sessions the speaker with the most desirable treble boost and bass thump (usually in the 100Hz range) was preferred at that moment. Much like the "retina burning setting" on TVs to sell in the TV store, a speaker with a thumping, but not boomy, bass and crispy, detailed treble almost always won out in side-by-side comparisons in the stores' listening rooms. So dutiful marketing departments and acquiescent engineers delivered. The trick was to please the comparison shoppers while still delivering great sound. that's how we got here.
During this time the audiophile speakers got more and more gimmicky - selling on technical concepts which sound right, but don't necessarily work in practice - like "infinite slope" crossovers, or the opposite goal of the "single pole" crossovers. People started developing magic tricks which often had nothing to do with the performance of the speaker, such as milling the flared ports of a bass cabinet out of 20 pound chunks of aluminum, or using a layered sandwich of MDF, sand, and MDF to somehow create a 3/4" panel for enclosures. Some of the gimmicky stuff actually make a difference, but in a large number of instances it is just nonsense.
I once auditioned a high end speaker from a well-established company where it was blatantly obvious that the tweeter was about 4dB to 6 dB too loud. When a reviewer asked the guy running the demo about it, he got a snotty answer. Later it was announced that the demo had the wrong prototype crossover and in future production units that issue wouldn't exist.
I totally understand the concept of using a different speaker to get a sound quality one likes to hear. That's fine. But, the concept of attempting to recreate a real accurate reproduction of a real recorded event is what most very serious music listening aficionados, like many of us on this forum, is what makes this a great hobby. If one wants to boost the treble, fine - just don't tell us it is accurate when we complain about it.
The problem with a great accurate speaker, as Rammis points out, is that alongside blowing us away with accurate reproduction, we also have to hear all the flaws of a bad recording. I will never forget Bryan's face when just after a serious listening session of a pair of Dynaudio C1s using my reference discs, he put on his beloved Radiohead CD and it fell very, very flat. Relative to the very clean and dynamic recordings we had been auditioning, Radiohead lacked any dynamics and the treble and bass had no real extension. He was so sad. But I say, so what. If the music moves you, it doesn't have to have the same perfect recording as something else which may sound fun, but doesn't give you life.