Aye, it is hazy...
But no, it's pretty clear...especially if you first look back, and then apply what we learn looking back going forward.
I should let you know I saw this future of the wi-throttle back between 2007 and 2008. It was poohed on because I was thinking in terms of the Nintendo DSI, and everybody could not fathom the DCC manufacturers producing the throttle, coding the software, and offering it on the market for anywhere near a reasonable price. These people, though, were in a tunnel that stated the controller HAD to be made by a DCC manufacturer and the programming HAD to be done by a DCC manufacturer. The technology I was invoking was way too advanced for our DCC manufacturers to afford, and they couldn't figure out how the physical button layout should look, and then ranted about how each generation would have to essentially be a whole new machine, so the DCC manufacturers would certainly NOT go down this road..
But then along came the iPhone later that year, followed a couple years late by the wi-throttle app...
I forsee that it will take as long for the present DCC population to start using a Smartphone instead of the Boondoggle, as it took for them to adopt DCC in the first place. This means some people will be dragged into it kicking and screaming, but once they use it long enough to get used to it, it will be over.
The kickers will be the same kickers that got them into DCC in the first place:
1, The wi-throttle technology will so quickly outpace the button throttles in terms of capability that it will be like comparing DC versus DCC in terms of operability. Refinement is merely a matter of programming, and programming is cheap [compared to making a physical object] for those who know how to do it.
3, The difference in price between the wi-throttle with software versus the button stick will be such that it make no $ense to buy the button stick. In otherwords, people will not be able to afford the command station, PR-3 and 8 button sticks so all their friends can come over to play, but they will be able to afford the command station and the PR-3, and then use the devices most people will already own at that point - which will be the ubiquitious smart phone.
2, The alternative will be either too expensive or simply not exist any more. A DT402D has a sticker price of $250 and it can only consist two locomotives at once. Wi-throttle, first gen iPhones are $120 on the high end on ebay and the application costs $0.99. And I can upgrade my software as it comes out. And as more people opt for the smart phone, less manufacturers will be making buttonsticks, while the buttonsticks will be priced in the clouds, and some manufacturers will drop their throttle product lines altogether...and/or go poof.
The beauty of the wi-throttle is that I can own one device and try out four different approaches to model train throttling and then choose the one I like best and use it - without a care in the world about what the layout operating system is [Digitrax, Lenz, NCE, Ring, Zimo, MTH DCS] so long as those other manufacturers support a JMRI computer interface. Those manufacturers who don't offer this support to app-phone interfacing, well, this is their stage right-left exit...they will be like RailCommand or Dynatrol or any of the other's here real soon.
Instead of picking out if I like Lenz's throttle idea, Digitrax's throttle idea, Lenz's throttle idea,Ring's RCC throttle idea, MTH's DCS throttle Idea...the list goes on..., I can just pick up my smart phone, upload all four, five, six...hundred? throttle programs, and try them each out in turn. I settle on the program I like, but if a new better program comes out tomorrow, i don't have to buy a new throttle; I just download the software.
I could then use program A while someone else in the same room could be using program B. Let's suppose I'm at Jim's layout, and Jim insists on using NCE system architecture but has Wi-throttle ability. We suppose I prefer Digitrax's GUI layout and functionality, while Bob prefers to use Lenz's GUI throttle app [neither exist yet, but these manufacturers had best be on the ball!]. I show up with my wi-throttle [the same one that guided me to his house with it's onboard GPS!], acquire his layout, and then run my train using the Digitrax program while Bob uses his NCE App on his wi-throttle. Jim still runs his trains with the Lenz button stick, but neither me nor Bob had to buy an NCE throttle to work the NCE layout, nor use any sort of work-around to use a different throttle on this layout, and Jim didn't have to buy three throttles so we could all play together.
When one program idea becomes obsolete due to a "better" throttle idea, instead of looking at buying a new throttle for $150 or whatever insane price they stick on their 1980's TV remote, I download the new throttle software for $0.99, or whatever it costs, and I'm on the road running; in 20 minutes, I either use it, or stick with the old one. When I upgrade phones here, I may indeed be joining the revolution, because I'd love to have the map features of a smart phone when I'm out and about. Hence, I'll already own the phone!
The throttle component, then, becomes is a null issue, and you're essentially hunting for the best power boosters, command stations and ancillary system architecture that supports the wireless-throttle approach.