What I've never fully understood is why there's only one SCORM. Why do people talk about the SCORM instead of a SCORM? The very notion that there can be a Shareable Content Object Reference Model for learning objects would seem to suggest that there can be more than one. What's relevant to the US Department of Defense in terms of e-learning is unlikely to be relevant to, say, training doctors. There's an a priori assumption made by many that one size fits all. This isn't a claim made by those developing SCORM but it has been an assertion that's formed nonetheless.
The work of IMS is to create interoperable specifications that allow domain-specific requirements to be catered for by selecting the most appropriate combination of specs. For example, a subject domain that requires groups of learners to interact can adopt aspects of Learning Design and Learner Information that fits their needs. A domain that's more focused on objective knowledge and the sharing of content might pay more attention to Meta-data and Digital Repositories Interoperability, and so on. More likely an individual subject domain would select a wide range of specifications to create a reference model and application profile. More precisely, they would create a SCORM for their domain (rather than adopting the SCORM that exists at present).
Expect to see at least one new reference model and application profile emerge over the next 12 months as initiatives such as IVIMEDS establish themselves.