Certainly an interesting development, and if I was a buyer, I would look at this as meaning at least another year. A few other thoughts . . .
As has been said, P-roc is a proven platform. Between BOP2.0, CCC, Predator, Matrix, Wooly, Buffy, Deadpin, the Kuglers, and plenty of others that have not been publicly available to play, it is clear that the P-roc works and is reliable. I will say that the driver boards have not been used as extensively at this point since games like CCC, BOP2.0 and Matrix were using the boards already existing in the game (Predator and TBL are both using the driver boards as well as most of the custom games (vs re-themes and extensions). Regardless those boards have seen plenty of action.
The FAST boards are new and have only seen action in a couple of games, and other then the EM conversion that was at expo, I don't think any other games have been playable at various pinball events, at least not that I have played (I think there is a portable machine that Aaron has shown with the boards). I am not saying the boards are not good or reliable, just not proven at this point. Their drivers boards have a little bit of a different 'philosophy' in combining drivers and switch detection into a single board, and only time will tell if that is a better, worse or equivalent as far as functionality and reliability (plus MPF will work with P-roc and associated driver boards as well)..
The pyprocgame framework is a proven commodity at this point, with all of the machines mentioned above using it. However, some of those machines are using extended versions of the framework that provide more support for color, larger displays and other features. CCC uses an extended version that Eric did to support color, DP has their version also to support color and their other enhancements, and Buffy and the Kuglers use what is called the HDVGA version of the framework, which is publicly available for others to use and has been shown to work with 32bit color running a 450x900 size display, that also has a new upgraded version on the horizon.
The MPF framework is full of potential and I think when it is finished (last I heard the v1 release was expected in about 5 months, but the alpha/beta is available now), should be an excellent platform and may turn out to be great for amateur programmers, since you should be able to get the basics functionality of a games mechanisms up and running with little to no actual programing. Of course the game won't have rules, sounds, graphics and all the rest, which will require programming, etc. In the case of helping JPOP, FAST and MPF certainly have guys with those skills. However, how much of the current code is usable with the new platform is unknown (at least to me), but given how MPF does things, there is certainly some refactoring needed.
I find it interesting that JPOP is now on his third (or more?) hardware platform. Based on Eric's comments it sounds like the game does currently run on the current hardware platform, and it is not clear why the change and hard to know if there is an issue with the platform, JPOP being JPOP, or if FAST/MPF made him an offer he couldn't refuse. BTW, multiple manufactures have looked at using P-roc and have chosen to go their own way, which I suspect is likely been under the belief they could find a cheaper solution as the P-roc with its computer requirement is not cheap, so maybe they did end up with a cheaper solution, but certainly seems they have had a lot of delays in the process.
Time will tell if this is this the next volume in the JPOP saga, or just another chapter. For those of us in the cheap seats, it continues to be an entertaining and enjoyable to watch. Hopefully at the end those who have bought will get their machines and laugh about the journey.
Did just hear about a study that found that the process and anticipation of buying something you desire, is actually more emotionally stimulating (typically positive) then when you actually get the item. So, for you buyers, this is all good