That’s not the case. If something’s better now, it doesn’t mean it wasn’t good before.
It's not the case that Cyngus is more accurate than ares or? Didn't it come about thanks to critical listening of Cliff, which revealed discrepancies? It seems that a big number of amp sims have changed now, some more so than others.
I mentioned accuracy, not being better or worse in terms of some other measure (for example "what someone prefers"), if it's about that. Of course older amp modelling can still be good or even great by one's taste. A lot of people like what kemper does to source tones, constant, slight TS hump. But it could be more accurate.
And for accuracy as well, if Cyngus is more accurate than Ares, that's cool for people who want that. But this does not mean it "wasn't at all accurate" before. It's not even clear to me at which point A can be said to be "accurate" and what exactly this entails (often a loaded description). The kind of test considered sufficient to call it a day and sims "accurate", in some companies, isn't the same elsewhere.
I do believe that most claims about fractal's innacuracies are due to bad testing or unwarranted conclusions of users. But also seeings so many amps change quite a bit is telling about how things haven't been "perfect".
Not that cliff has claimed this about all amp sims (I don't know if this has occured about certain sims). And I like the fact that fractal keeps moving forward instead of just calling it a day. It's such a different approach from some other companies and definitely welcome for me personally.