Cough*Utility Data Center*cough
Whilst HP's approach is applaudable, this is not the case of the wise, pragmatic tortoise, but more of the experienced hare that got kicked in the family tiara jewels, and has learned from the experience that sometimes it's better to jog rather than sprint. HP originally did have a very futuristic, ahead-of-the-market, big-bang datacenter product that was a complete computing-as-a-service offering, that could be built by a customer to convert their whole infrastructure in one go to a billable model using smart servers, storage and software. It was the HP Utility Data Center and it even worked!
The problem was, apart from three big customers, everyone else couldn't afford to rip out all their old datacenters and replace them overnight with shiney new ones, they wanted a phased approach (Ashlee Vance wrote about its demise here http://www.theregister.co.uk/2004/09/30/hp_kills_udc/). So HP canned the offering, remembered how it had originally found success with the OpenView products, and broke the Utiltiy Data Center up into standalone modules and services. Those bits sold well, and HP haven't forgotten the lesson. Business usually transitions slowly to new ideas, it takes legislation (e.g., Sarbanes-Oxley for retention) or a dramatic turn in World economic events (9/11 for DR) to force the pace of change to pick up, and HP have decided the current downturn isn't sharp enough (yet) to force the wholesale adoption of rent-a-cloud services.
Long may the hare limp!