To date cloud computing and its offshoots have been described in zero sum terms: Amazon's storage system over another vendor; Salesforce.com's platform over Siebel; IBM's cloud over HP's. The reality is likely to be much more democratic.
Vendors offering specific services, such as cloud-based storage, will not dominate, unless they can team up with other partners in new cloud ecosystems.
In a nutshell, the cloud will be populated by numerous vendors--SAP, Oracle, IBM, Microsoft and others perhaps companies like Amazon. Sounds a lot like the current enterprise infrastructure. The difference is that these various cloud vendors will be glued together.
The cloud ecosystem theory builds off a Bernstein report outlining the landscape. In the report, Bernstein analyst Jeffrey Lindsay argued that Amazon's cloud efforts wouldn't be a massive hit--at least relative to the etailing business.
As far as vendors go the winning cloud players are either going to have the actual data centers and hardware or the software that helps manage the ecosystem. If you buy into the ecosystem argument it becomes readily apparent that cloud computing is in the early innings of its development. While many companies have cloud computing efforts few are talking about slapping all of these clouds together or managing the various parts.