Errrr.... the amp modeling of the Axe Fx ---IS--- software!!!! The sound of the Axe is created entirely by software algorithms, nothing else. You can run these algorithms on any platform.
There are only two things to consider to port Quantum modeling over to a VST/AU/AAX Version:
- The computer you want to run it on needs enough horsepower. That could still be a problem, I don't know how the current line of Intel CPUs performs in such specialized DSP, but this will certainly not be a problem anymore in 3 to 5 years with the permanent increase in processing power.
- You need a good input stage so that the Quantum algorithms have a good signal to work on. I have made the test, I recorded some dry guitar tracks first through the 500 kOhm DI of my Steinberg UR28M audio interface, then through the 2.2 MOhm of a Universal Audio 2-610 preamp into the line in of the UR28M, and then through the Axe Fx II straight to the DAW via USB without processing. Then I reamped all three via SPDIF through the Axe Fx II with the same patches. Result: Axe Fx II of course the usual excellent quality. UR28M DI sounded WAY worse than directly through the Axe Fx II, the highs were rolled off, the life and detail was missing, night and day difference. UA 2-610 DI sounded awesome, it had even a bit more sparkle and life than the Axe Fx II direct tracks - not much, but noticable. So, a good input stage plays a major role, but is not something you couldn't get without the Axe Fx.
So, given enough CPU horse power and a decent DI, there is no reason why you couldn't get Axe Fx II Quantum quality from a plugin. The reason why all the Software modelers are worse than the Axe is because the algorithms are worse. Cliff is a genius in developing them, that's all (and that's a lot!!!). The Axe Fx II has no "magic hardware", it has just enough DSP power and high quality signal path, that's all. That's not said to belittle the quality of the hardware - but the hardware is just a platform for the algorithms, and if they suck, the tone sucks.