Rory,
my sweet dear friend and comrade in arms (and pain?

)!
I think, the first thing to do is to come up with a template for NEW call centers.
I am thinking about a simple matrix listing the equipment necessary vs. the functionality and size.
Here is a thought:
break the call centers into categories such as inbound, outbound and blended.
Then for every category, create different sizes: 50, 100, 200, 500 and above.
Ok, the easy part is over!
Next step is to get everyone agree to limit the scope of the hardware to be used! I am already hearing a collective "what the f*ck!". But, let's settle down for a second and think about this:
why is all the flexibility necessary? To be able to squeeze Genesys into any environment you might ever consider. This is good for Genesys, but then again, they do not have to do all the quoting all the time. I think it is time to stop advertising Genesys and put it where it belongs into the box and away from sight. (I think Genesys sales will kill me for this).
Let me compare Genesys to Windows XP :
new company is started and they need an accounting package. Your accounting package (PeachTree?

) is running on Windows XP. Do you care about all the different layers implemented into Windows architecture? Yes, kind of. But it is not important to the client. They need to do accounting so why should they worry about WinSock in Windows XP or how the keyboard driver operates? All they see is: PC and PeachTree screen that allows them to enter data.
And Genesys is not even Windows XP in that example but one tiny function.
So, the point is:
call center needs functionality that is very abstract: call logging, voice recognition, route calls to agents, see reporting, and so on.
So, why not just focus on the abstraction layer ? Forget about all the different variants of PBX and CTI and so on and let's focus on the main three or four possible choices? Customization means "varying from default".
Standardization is the main reason for Genesys success. Unfortunately, Genesys CTI is standardized only from inside. So, let's take the next logical step and standardize the whole offering? No more solutions. People get a box with all the necessary stuff inside that can do MOST of the things they need. AND THEN, if they want, they can customize it. Would it not be better for all of us?
The easiest way to focus on the abstract functionality is to limit the hardware offering. I know a lot of vendors would not be happy, but, guys, we all know that 99.9% of the functionality can be offered by whicever product and we are spending 99& of our time on the last 0.1%, which if you think about it can be accomplished with any product out there anyway. I do not mean to start an ugly argument here. If did, please ignore it, for it was not my goal.
I just think we need to stop using three hundred different types of PBXs and IVRs and so on just to fulfill the last 0.1%. And I want to do it in order to make life easier for both us and our enduser. We need to learn how to say "No, it cannot be done" because in most cases there is a reason why this functionality is not there.
So, without further adieu, how about defining PBXs, IVRs and voice recorders to be user for different types of call centers and then seeing if we can limit it even more?
For PBXs, how about sticking with the biggest three or four for all the new call centers? Avaya, Nortel, NEC and Rockwell? IVR: BrookTrout and CSS? Logger: Nice?
Any thoughts?