Knowing how a Kemper works, roughly, vs. the Axe FX, and now knowing that Cliff has something up his sleeve, I was going to say that I trust Cliff to be able to be more thorough and accurate than a Kemper's "one size fits all" profiling algorithms. The thing that always sold me on the Axe (and, honestly, I think also gives the Helix a leg up on Kemper) is that the Axe is using component level modeling with algorithms for each component. Add on the stupid amount of processing power and Cliff's ability to "profile" an amp component by component down to the exact tone stack response, I would take the Axe any day. With the Axe I know that every setting I do to it is going to react just like the real amp would, whereas the Kemper runs into limits based on the quality of the profile and the limitations of its type of modeling.
Would I love to be able to have amps like a Fryette Pitbull or Fryette Memphis, or a Bogner Goldfinger? Sure I would, because I own two of those three amps and they're unique and amazing. But I can get pretty close with what's available already by tweaking the amp block settings. I already "mod" the Plexi models to suite my tastes with the saturation control and bright cap (honestly, the bright cap setting is, I think, one of the absolute best tweaks available to us, it's incredibly powerful), and tend to throw EL34s in the Euro models since I don't like the 6L6 versions of the XTC.
My general sense is that if you can't create something amazing using the Axe FX (like MANY artists already are), you're doing it wrong. It's about the music, and the vast majority of listeners aren't going to care if you used a Plexi or a Goldfinger and probably can't tell the difference on a recording anyway. They just care if your music sounds good.