if someone understands why using a tube power amp, may not be "tonally accurate" but still wants to use a tube power amp because they like the sound better, how is that incorrect? How can it even be less than ideal? It's their equipment and they're using it for their enjoyment.
I never said, or meant to suggest, it was incorrect in that circumstance (not sure who you think is saying that). I think it's incorrect and less than ideal if the goal is to reproduce an amp model as closely as possible to the modelled origional, but the coloration of a tube amp used for monitoring is taking the model in the opposite (to accurate) direction. On the other hand, if the goal is enjoyment of the tonal nuance that the tube amp adds which the user happens to like, then ya, totally correct. Depends on what the user wants to do. Some are sold on tube amps and they don't know exactly why - my point above is that, for the "sold on real tubes group" it's all good and correct if they like what they hear, or if the output matches what they need or perceive as a representation of a given real world amp,
however, if they are, for example, plugging a Vox preamp model, or Vox pre+p.a. model into the fx return of a recto head + out to a Mesa412, and are struggling with their Vox tone and asking for advice, then the conversation can be challenging if they don't want to let go of that recto which is, yup, incorrect as a solution to match their use case. In such cases, where the user really likes real tubes for whatever reason (or for no particular reason), but struggling with some modelled tones accuracy, then I'd recommend, as a logical approach to get as much modelling accuracy as possible while retaining a real tube output for monitoring, that the user look into a hi-headroom / low (ish) coloration tube amp like LxII or similar, with Axfx p.a. modelling ON depending on the model (because LxII will not likely be able to produce on its own (with Axfx p.a. modelling off), even at higher volumes, the pronounced p.a. character of many amp models with p.a. modelling ON - (nor do we want it to in order for it to be viable across varied models), but given its hi-headroom, lower color'd character, LxII or similar, should be able to do a reasonble job rendering the active p.a. aspects of the model), and IC modelling set flat if using a regular cab (because a tube amp + regular cab will produce that automatically). Will it be as accurate as with a SS amp just playing a "make louder role?, probably no imo given some wildcard variables introduced on top of the modelling by those real tubes, but likely in the ballpark of accuracy, and much better than using the traditional amp head return. + the user gets to stick with their beloved real tube nuance (which by the way, I'm not convinced is just perception, and, despite being real close, I'm not sure is fully captured in the box yet - so far confirmed by continued regular updates that everyone seems agree gets us "another step closer" to the real deal).
It's a similar discussion wrt real cab vs FR cab / FRFR, and mic'd cab vs AITR. But when specific needs are stated, as in the examples I made above, I really don't see at all how the solutions are just subjective and thinking thru a logical process is not involved. Of course it is. So I'll continue to be irritating to the "J.U.Y.E." crowd and avocating for EEB (ears + eyes + brain).
Edit: Part of the issue here, in many help-me theads (not necessarily this one), is that the actual needs of the user are often not that clearly established, so you have all kinds of approaches from JUYE to "dood, you need an xyz" being shouted out from all directions without much context (if you have the $, buy some CC or LT or ... consultation time)