JBL Studio 4410, Wharfedale Elysian 4, JBL Array 1400, JBL S312, JBL S38, Klipsch RB1, JBL S412P, JBL L100 (European/Denmark model)
To quote my sister "this does not sound that good" then, connecting Pioneer SA9800 and saying "ok, that's it, it's back!"
Without any confirmation that any of the amps were still performing up to design (at a minimum) this is just than another anecdote with no proof. Did you measure each amp to ensure proper operation as in measuring it for gain, linearity, distortion, power response? Every channel? Under intended load? Were bias settings confirmed to be set per design? I say that because I'm using Kenwood KM-X1 6-channels amps manufactured from 1993-on in my main system. I can confirm, from my own measurements and setting bias myself according to the manufacturer's maintenance manual that any amp, especially old ones but even newer ones, very often have drift that does change the sound because the amp is not performing to specs. That drift could be thought to be "better" simply due to non-linearity accentuating a frequency range that gives the impression of "better".
As my own anecdote, I used a 2-way design of my own that I use with my Aragon 4004 II at times. I had recently purchased a Fosi V3 Mono to use in a distortion measurement rig and confirmed, through measurements, that it's performing close to spec, certainly well within audibility. As a test I substituted it into one channel of my 2-way setup, so one channel Fosi, the other Aragon 4004. Guess what? No discernable difference. That 4004 I purchased new in 1992. If my Aragon died, I would not replace it with a new Class AB amp of any make, I would buy one of any number of newer Class D amps covered in ASR. Is manufacturer quality and responsiveness important? You betcha, but that and functionality (pure amp, no built in preamp, no volume control, etc.) would be the deciding factors. And price of course, but that's another topic. No more Class AB amps for me, reality is that as long as the amp has the power capability for the intended usage (don't need 400W for my desktop system) and that it can handle the load (I have no 2-ohm minimum speakers), there's little chance of any audible difference.
But that is all anecdote. I can't prove any of it. Well, I can and actually have proven the Fosi amp response in another thread. That said, I also purchased another pair of V3 Monos some time later for desktop use. I measured the response of both in my distortion measurement rig. Both amps had serious distortion issues, so I returned them to Amazon. The linearity was actually fine, but the distortion was well off that of my first one. It might have been possible that I would have heard a difference in use, but when I saw measured distortion much worse, they went back. Had those been used in a "comparison" for sound it well might have been that they sounded "worse". Not due to design, but possibly due to quality controls, etc. That happens with other amps as well, Class AB or others. So it's necessary to be able to prove, at a minimum, that any component in a system is up to specs if you want to prove audibililty of anything. Speakers, too.
I guess it should go without saying that I measure everything now that I can. I've been surprised at some underperformance I've found as a result.
That's a minium to start. I can also say that I would never accept my wife's impression of something being "better". Different, possibly (or not in actuality). Proof of any inferiority or superiority? No chance.
Edit: Also had to replace all the relays. They (KM-X1) failed or were failing with questionable contact condition. Just another possible issue in an amp.
) elsewhere - or at the very least comply with