As much as I wish your point was valid and a possible future, I don't think it is even in the realm of possibility
At least with analog gear
That's already basically the direction modelers are going, although the available features in that vein don't do it that well IME
The two notes wall of sound is basically different EQ filters to simulate 6l6 vs el34, etc, because that's what digital is programmed to do right now
I mean, listen to the "great tones" in popular media - you can't convince me Tammy Henson or the Periphery street shitter or YouTube Shill Number 83639-26 aren't widely lauded as Good Tone™ because of the creeping homogenization and unimaginative tone seeking modelers have contributed to
And I honestly don't think it's Kemper or Fractal or whoever's fault, and I honestly don't think they sound bad or are bad products
It's because they promote the most banal, unimaginative, boring tone-seeking from the popular guitarists that use them
I mean, I think the one caveat to my point of view about modeling being a force for good overall, and that it could even have an even more promising future not just in convenience, but in tonal innovation, is honestly simply because Fractal Audio is part of the discussion. I think in a world without Fractal, I'd be right there with you. But dude the Fractal stuff is so good and allows for so much dialing in that I'm excited for the future of modeling almost specifically because of what that company is doing.
Even with Fractal's stuff,
right now, today, you can get crazy deep into modding any given digital amp into something else almost entirely. You can't quite get to the level of swapping individual caps and resistors anywhere you want (only in some places), but you can do crazy things no layman could even dream of doing analog. All it would take is the right people doing modeling to move just a little bit more in that direction. I don't think it's as big of a leap as it sounds. Especially if it's treated as some additive feature you're not forced to engage with.
I do agree with you that most people who play guitar probably won't innovate and just want the easy way out of getting "good tone" by copying other people's sounds whole hog without thinking about it, even downloading other people's presets for tone, because they can't even be bothered to learn the tools they're using. And I hear what you're saying about innovation in the analog world. Getting good guitar tone with analog gear is fucking hard and quite complex, and even if you know exactly what you're doing, you won't really be able to exactly duplicate other peoples' sounds, which will lead to all kinds of different tones and then different people chasing even more different tones still.
But I still think innovation is just as possible in the digital realm. Yes, lazy people will use other people's presets, but if the future of modeling leans into making it ever easier for more non-electrical engineers to express themselves by creating entirely new sounds in ways the world's current tube amp and pedal designers haven't allowed for yet, I don't see that as a bad thing.
I also don't know if I can squarely blame modeling for any homogenization the guitar community might be seeing. I think that has more to do with how accessible home recording and just doing everything yourself has become over the past decade or two with consumer recording electronics getting ever cheaper, in combination with social media being right there for everyone to show all their tips and tricks on.
Modeling didn't cause every heavy player to gravitate to a V30 4x12 -> SM57 setup, pros showing everybody how good that combo sounds on social media did.