Ho how i hope answers will not be '...but VCD is so easy to use ..
Well, it's more than that. It's "ready to use". 
What you described in your post is a Boeing 777..... in 500 pieces on the floor. It doesn't fly.
Let's say that vCD (1.0!) is a Cessna... ready to fly, you can turn the key and you are ready to fly.
In your analysis (which I am not arguing with) you are missing a key point. To reach the ease of use, integration and user-experience that vCD provides out of the box you probably need to spend a few M$ of cash + a few months / years of customization.
That's the value of it. Not necessarily the richness of futures of a point in time technology (one of 500 comprising the whole solution).
Having this said, as I stated somewhere else, I can imagine vCD evolving over time onto a path were we add more features to our pieces as well as we allow easier integrations of third party features. This is my opinion and not an official VMware statement.
I don't understand your parallel with open-vz manager. I am not suggesting to use an inferior and cheaper technology compared to a better and more expensive technology. I am however suggesting that there are good reasons in using something when its overall value is much greater than the sum of the values of the standalone parts that comprise it.
Massimo.
Massimo Re Ferre'
VMware vCloud Architect
twitter.com/mreferre
ok, both points of view are clear, and i like the allegories you use 
now, it is also clear that a low-end alternative to services might fly for SMB, commercial, as usual, but not for ENT or SP class IMHO. (when i say 'fly' i don't mean they won't buy the VCD licenses and use it, i am saying end-customers will not migrate to cloud if infrastracture services will not be the same, exactly like they did not except virtual server alternative before vmware prooved VM provide the same capabilities as physical server and some more).
what i see coming (and several cloud analysts and surveys suggest -num#1 inhibitor for cloud is security, num#2 is support for legacy DC components) is that at first phase your cloud customers will want to migrate current 'hosting enviroment' into 'the cloud' and this will include common services, migration of features will simply be impossible so they will need to look for solutions. A 'grienfield' SP might be using compute-only cloud which is totaly fine with VCD, for the long run customers will need to bet on the boeing 777 because it will be the one providing real value (and several boeing 777s are coming into the airfield soon).
Future a side, several things must be fixed ASAP on VSE in VCD before going production : stateful failover (absolutely crucial if you want cloud to have same SLA as common DC from 10 years ago) , VSE per ORG rather then VSE per network , VSE with more then 2 vNIC , workaround to BW issues with VSE (vshield APP, zones instead) and of course routing support (instead of just NAT), all of which and others i have posted here as field engineering feedback.