I guess I don't see the point to the all in one device. So what if I can use the same protocol from my HA system for each device technology, I still need a HA system controller or PC with software to handle the logic and end user GUI right!?!
In reality, I need a PC to stream content throughout my home anyway and want a touch screen interface that I can easily modify, so why not use free home automation software that works with each technology already, but runs on my PC?
Also, PC's are reliable. Most folks may not know it, but many critical control systems in several industries use XP! Of course, such systems have fail safe designs so as not to cause an adverse condition if the PC fails, but I'm just sayin' if a PC is good enough for multi billion dollar industries why can't XP embedded work for me in my $200,000 home?!? Further, since most PC based systems are made up of smaller sub systems (security, HVAC, lighting etc...) everything still functions if the PC fails.
All any PC software has to do is use a generic and robust scripting language and have all home specific code separated from the device specific protocol that handles the actual commands and it has duplicated the function of black box in a much more versatile way. This was already done with Premise many years ago and it still works very well. If it takes that many more hours or days to set up a home over hard coding protocol commands, I'd seriously question the PC software you are trying to use because it makes much more sense to let a PC manage the device specific protocol and give an integrator a higher level generic scripting language to manipulate the device. I know Premise cost millions to develop and that Motorola now gives it away for free, but the designers were top notch programmers from Microsoft, so the program is 100% stable under a windows environment (and mature for that matter as they spent many many years perfecting it before Motorola gave them the pink slip).
An example on versatility:
I can change every light in my home from z-wave to any other lighting technology and retain all of my room specific logic (e.g. occupancy timeouts, light dims when I hit play etc...). I can do this in under half an hour (plus the hardware change out time). If you've hard coded protocol commands into a Creston or AMX system, not only would it take way more set up than half an hour, it would be much more error prone. Yes, I know this new black box could do this too, but it costs money and is not needed since PC's are becoming very cheap and most already have a PC in their home.
Another example of versatility:
If I wanted to install a system for someone, I can create generic templates using Premise. The templates give a starting point on basic logic, rooms and home devices. They can even include logic to brighten a room when a PVR is paused, occupancy time outs etc... Next, since all device specific code is separate from any customer specific scripts, I can easily import any device modules I've made for previous device types or download one's others have made. Next all is needed is to "bind" the device object to a home objects. Since home objects (lights, door locks, tvs, etc...) are separate from device objects (z-wave dimmers, X10 sensors, Elk M1g, etc...) all home specific logic can be retained if the device object is deleted and replaced with one of a different type! Yes this new black box could do this too, but again what I'm talking about is free assuming a user already has a PC.
My last example of versatility, what if a new device protocol comes along and the black box doesn't get an update?:
If a new device comes along it may never be supported by the black box, but I know for sure I can code a new device driver for Premise very easily! Good examples of this are the Elk M1g, VRC0P (Z-Wave), and W800RF32. All these drivers were created after Premise was no longer supported and were even created in vbscript (an easy to use object oriented language)! Best of all the drivers are free as most users publish their work, so others do not have to recreate the wheel. I'm not saying versatility ins't possible with a black box, I'm just saying that if it's not, I'm never going to buy one!
One could create some of this in the form of a black box, but without a robust object oriented scripting language, I seriously doubt it would be user friendly to ever add new devices. The problem is making a system that does this takes years, lots of talent and money. Please don't take this the wrong way, I'm all for start-ups, but I figured I'd throw out my thoughts since you asked for our opinions. Personally, I'd run from trying to penetrate an overly saturated home automation market with a small user base, but that is your choice.