Albert totally nails it in his post on the Learning Objects "debate". Basically - get over it. Move along. Do (and use) whatever is appropriate to what you're trying to do. One size does not fit all. Caveat emptor, etc...
I especially like his tips for subscribers to "information transfer" vs. "social constructivitistic" paradigms of learning objects (and, I would suggest, of learning in general).
But wait! There's more! Albert offers a website/wiki on "virtual apparatus", which appears to be a set of guidelines for creating content in a consistent manner (did I interpret that right?).
We've been talking about "learning objects" a lot around the software side of the Learning Commons over the last few weeks. We've been sort of stuck in discussion, tripping over the concept of "reusability" - there's the technical reusability (interoperability, via things like IEEE LOM, SCORM etc...) - and then there's pedagogical reusability (dealing with the content itself, and not the transport/interchange format used to squirt it over the 'net).
In our discussions, we're trying to plan out what we want to accomplish over the next year or so - what software needs to be built to achieve the goals we set (once we set the goals, of course). It's been interesting, if frustrating, but I think we're making progress as we start to realize we're all talking about essentially the same thing - just coming at it from different angles (pedagogical vs. technical reusability).
So, there's a pretty solid example that the definition of "learning objects" - and the implications that come with it - can be confusing or misleading even within a small team. No wonder there hasn't been any real consensus in the education community as a whole...