This is a very interesting discussion. On the one hand, it is of course ridiculous to judge a whole format on such grounds, as the circuits inside a large number of euro modules are similar to those in 4U or 5U, and nobody could probably hear any quality difference in a blind listening test between formats. Euro is simply full of amazing circuits that sounds amazing, with lots of sonic depth, fantastic timbre and heaps of character. On the other hand, I understand exactly what you mean, if you think of timbre as an expanded concept of user experience. And I should clarify that I too come from more than two decades of using other formats (Wiard, Blacet, CGS/Serge, Buchla, Bugbrand and other bananafrac, etc, and still using them), but started a euro system on the side a few years ago to have access to more complex modules that are not available elsewhere.
You can look at timbre as an expanded concept, not just including the spectral and textural qualities of a sound (and their temporal development), but also including the feel of producing the sound, the experience of playing it (including the feel of the knobs, the jacks, etc), the associations and memories that are included in making them, their musical context, and even the gestural movements required to play it (sounds requiring related gestural movements, or emerging from similar objects/modules, or which relates to some similar musical memory could be (i.e., perceived to be) timbrally related if using this expanded concept of timbre, regardless of sonic similarity).
If you think of timbre like this, which we may do more than we are aware of, as we are affected by our experiences when evaluating what comes out of the speakers when we play, the experience of playing eurorack could sometimes be perceived as "shallow", "small", "flimsy", "lacking character", etc. It would not be detectable in a blind listening test, but since these experiences from interaction while playing potentially affects me as a player when I make music - as part of a musician-instrument ecosystem - it also affects what and how I play. Hence it implicitly makes a musical difference.
If you want to read more about this expanded concept of timbre and its implications for music-making, from a musician's perspective, I can recommend the artistic research doctoral thesis from 2020 by wonderful Berlin-based improviser Magda Mayas (of which I was the main supervisor, so yes I am biased). She's an experimental inside piano player, but it all applies to electronic musicianship too:
https://www.researchcatalogue.net/view/382024/382025
Getting back to the Generate 3: Another interesting observation is that some people are interested in the sound - like looking for that perfect violin. Others are interested in variability, as in new instruments with new ways to control and play sound, or new ways of abusing a violin. Myself, I am certainly of the latter kind. From that perspective, regardless of what you think of the sonic timbre space of Generate3 (which is vast), it does indeed provide a unique parametrization of the timbre space, which simply cannot be found elsewhere. That is, it gives us new axes along which we can explore the sonic possibilities. The knobs and modulation inputs simply control different aspects of the sound than in most (or all) other oscillators. To me, that alone makes it super interesting as a tool for sonic exploration. And engineering-wise, I am still amazed that Joranalogue has been able to create such a novel design - I am still wondering how the circuits are even possible. Some serious analogue tricks going on inside.
To me, Joranalogue designs are very similar to Serge in their atomic design of interaction, control and patch-programmability (and I am not the first to make that comparison). I'm sure the circuits are very different to Serge, but the interaction design approach is somewhat similar. Focus is on variability and versatility. The modules are split up into basic functions, and you have to patch/modulate it to sound fat or dirty or deep, if that is what you want. Or weird or clean, if that is your preference. You can abuse this "violin" in so many ways. In the other end of the spectrum is, for example, to take another favorite brand of mine, AJH, who makes modules that have less parameters, but a special character already built-in - a Stradivarious of "violins" (or VCOs, or whatever kind of module you are talking about).
I regard Joranalogue (together with Frap Tools, AJH and a few others) as one of the most sophisticated and interesting of the current generation of analogue module designers. And of course it is very personal, but I find it quite a bit unfair to say that their modules lack character or depth - as the character and depth of the interaction potentials are simply both novel and vast.
All of this IMHO, of course.