The OpenStack effort is aimed at setting the "rules of the road" for the majority of cloud computing interactions. In this Forbes article, the author describes how the best Open Source projects are generally a meritocracy; a philosophy that promotes the notion that power should be vested in individuals according to merit. Advancement in such a system is based upon perceived intellectual talent measured through examination and/or demonstrated achievement within the community. In the case of OpenStack, this is demonstrated by those who contribute code with a variety of companies currently vying for the title of top contributor.
The OpenStack approach ensures portability. Public providers (cloud service providers anyone can use) should adopt the standards that the "body politic" endorse. More customers will come to you if you can show how easy it is to migrate. Of course, all the 'big boys' don't want to make it easy to leave their service, the flip-side of such ease-of-use.
From an application development strategy adopting a service oriented architecture is a "from the ground up" approach to making portability work. As long as end points are web services, theoretically one can build solutions in *any* IAAS/PAAS. Linked via an enterprise service bus (ESB), for example, data and business logic could managed with the ultimate in reliability: having the components spread among many cloud providers. One could always move services around (assuming the provider supports the native service language (dot-net, J2EE, etc.) or data repository. With the right planning, data could be independent of the data store itself (remember XML databases, or that old standby standard, SQL92?).
Utilizing standards and open source software will aid cloud computing to achieve the kind of market success seen with the World Wide Web, e-mail and other widely used Internet offerings. And that's good for everyone.