Fwiw, in general, I'm wondering whether capturing in itself will one day become obsolete.
The top tier modelers are already covering sooo much tonal ground that pretty much everybody should find a suitable amp for pretty much any situation. And by now even the cheaper range of modelers is massively catching up in terms of quality and quantity.
Obviously, the main advance of component modeling being that the model is in fact acting like a real amp, something just not possible with capturing (as has been discussed ad nauseum), even if advanced techniques such as Kemper's Liquid Profiling might close that gap to some extent.
IMO, captuing as is might only stay relevant for some folks with whatever unusual amps (be it rare ones, modified ones or whatever) and/or combinations of amps and stomp boxes. Anything more typical will be covered sufficiently by component modeling pretty soon, very affordable options included.
Capturing IMO would only be able to stay around permanently in case it'd really capture pretty much everything I'd throw at it. Let's say I'd be splitting my input signal by frequency and then amplifying both splits quite differently (which could make a lot of sense in some situations, such as in keeping your lows tight and clean while getting successively more driven the higher you go (I know, sort of like the old TS808 trick, but I'm thinking about a more drastical approach). From all I know, such creations can't be captured very well right now.
Another option for capturing to survive would be if there were ways to represent all (or most) of the captured amp within one single capture, possibly by utilizing some clever interpolation algorithms, along with certain refining procedures (perhaps that will never really work, but getting a lot closer should be possible). But until we'll be seeing that, I'm sure the market is flooded with component modeling offering pretty much any amps that ever existed.
Yet another thing would be if capturing could profile even some truly weird sounds. Such as the tones of an SY-300/1000 (obviously in plain pickup mode). Or some other wild filtering creatures and what not. But I don't see that to happen any day soon (btw, would love to know what, say, NAM makes from some SY tones...).
Personally, the only reason for me to still sort of lust after a Kemper is it's IMO excellent onboard UI and the established libary. But the latter will become less and less important. Fwiw, at first I thought I would defenitely buy a ToneX pedal one day - but right now, I wouldn't even know why, especially as I'm generally fine anyway and as the HX Stomp has seen two massive updates which should keep my tonal needs sorted pretty much forever.
Really just wondering how much of a market there still is for capturing devices.