You want data? Good luck.
Audio measurements that would reveal how close or how far an audio device duplicates or approximates what we hear remain stuck in a primitive state.
For one, how would we correlate what we hear to graphs or numbers?
Who has the authority to establish what specific pattern on a frequency response graph (which is usually derived from a sweep of a signal generator-produced tone, played back over a supposedly calibrated loudspeaker) is closer or further away from how we perceive sound?
To use an analogy from another of our senses, the visual: how would we quantify the specific plasticity of a photo made with a Leica M3, compared to the same image rendered by a Nikon F2? We cannot, because, lacking objective criteria, we must use subjective interpretations. And in the realm of the senses nothing but these interpretations matter, because they all end up as sensual perceptions, not verifiable, desirable or objectively interpretable data.
In the case of VF14 emulations, both tube and solid state, we don't even have to go that far and search for interpretations of reality* we may like more or less: they all suck, even on a primitive, measurable level: unacceptable noise floor, gross aberrations in frequency response, etc. Then you have the odd dynamic response common to all solid-state versions. Roughly the same as with all complex solid state processors of sound: the best of them ("best" here used in relative terms) make you feel slightly removed from the music, a veil over the sound that is not easily quantified, but instantly qualified.
The business end drives these products: money for nothing and the chicks for free: at best you get what you paid for, and often much less.
So, storing or keeping a backup to your Porsche engine in your Porsche inevitably points to a Porsche engine...
* for more on "reality" in microphones, read here:
https://repforums.prosoundweb.com/index.php/topic,37170.0.html