Ok, didn't realize you were looking to get into the integrator biz if this works out, that changes things completely, and there are a few new points I haven't made. My recommendation is that if you're really interested in getting into this for $$ is to think past installing your own system, assume that's a done thing regardless of package, and think about what you'll do once you start selling it. (btw, some folks are running CQC on a 40W low power mini-itx PC, to make it more "controller" like).
1) According to some of the pro's that I talk to on a regular basis, VR is not yet ready for problem-free deployment in customer's homes without massive tweaking, so I wouldn't use that as the tipping point for anything. If you think installation of CQC might be too much, you'll be crushed by the tweaking required for VR. (if you want details, ask QQQ on avsforum and remotecentral; he's played with it for years without acheiving the desired level of success for professional implementations)
2)
If the ease of install changes and both are pretty equal on installing and changing, I think my focus will then be on what system is more flexible and stable.
With all due respect, I think you've got that backwards for a pro install. Certainly for a DIY'er, they need rapid deployment because they're not making $$ at this. Plus they can deal with resetting stuff here/there, and a lack of flexibility. However, a pro needs to know that the tech support won't kill them, that they can easily modify to suit divergent customer types. You can come up to speed on any system incl CQC in not much time at all, and once your base templates are built out you'll roll them out to anybody who buys that system. That's how pretty much all the pro's that I know of work regardless of CQC/AMX/Crestron/etc; while they may tweak certain flows for clients, they rarely diverge greatly from the base template set as that's how you introduce new bugs.
I'm not saying anything about the other packages, only that at least for the very successful integrators I talk to, they've got 100+Hours in their templates, and for them flexibility to create the interface of their dreams is paramount. Like you, they only want to support 1 system, so ramp-up time is less meaningful than deployment and customization time.
3)
The choices have been made at this time on the hardware and most of the wiring is done....One reason is i dont see alot of my future customers being intersted in running a bunch of hardware...I was a vendor I would bend over backwards
Again, think past your own install given your future desire. You may have decided on your hardware and can now do what BSR suggests and see which package works better with that, but that doesn't mean one of your customers is willing to buy everything that you did. They may have a buddy who works for (Lutron/Autopatch/etc) and recommends it strongly, or heard good things about it, or have a large investment in it already (either current home or other), or their buddy can get them a deep discount. There's any of a gazillion reasons they may not want *all* the same hardware you do. Matter of fact, talking to the pro's, they rarely want the same shopping list.
In that case, support of a high # of hardware devices, and ease of integrating new devices, is king. CQC supported 100 devices on 8/2/06, 125 on 1/29/07, 140 on 5/9, and of the
49 in progress now, there's already been several more submitted for V2.2. Given that CQC distributes the driver development tools to everyone as part of the license with no NDA required, it's enjoyed a rapid rise. Plus they're all written in a CQC native langauge (either CML or PDL) so it's a "contained coding freedom" and there's no possibility of a badly written or rogue driver bringing down your system. There's 23 different users or integrators working on those drivers, and these have a history of not getting completely done until there's a callout in advance of a release, so I wouldn't be surprised to see many of those get submitted.
4) If you really want "HA model home" that will drive buying decisions, don't forget custom template PDA support. For my wealthier friends, that's the "flashy" stuff that makes them drool and want to pull out their checkbooks. When I told them I could use pretty and intuitive screens to adjust my HVAC/security/etc over the Cingular network, they were shocked. See if you can replicate this screen in the other packages. I know of at least one where it's physically impossible. No idea if that's due to an architectural shortcoming or some other reason.
Plus, it's not like you need to learn another language for this stuff. Use the same interface editor that you do with your other templates, just make the template size smaller and steer clear of the widgets not currently supported by the Dot Net /PDA viewer. (only a few of those).
These screens literally took me <30mins as all I did was copy/paste from my existing setup and resize the objects.
The "expert" screen that only I use (too complex for wife). The buttons are functional - I can turn on/off the lights, plus pressing the "disarm" button pulls up the security screen. During the heatwave in the summer, I used the HVAC one a bit to adjust the fan and cool down the house a bit before I came home. In the winter I turn up the heat when i'm enroute home, rather than leave it on some form of timer. My life is too random for that.