Hi Dai,
as I understand (and I must recognise that, currently, it isn't too much), LD defines some "artefacts" (roles, activities, resources, conditions...) to be able to construct learning scenarios, with paths, activities... It's really a cool step beyond current "content-based" specifications, sure. That's a great difference and, in my opinion, one of the bases of LD current acceptation from a lot of tools, educators, developers... Congratulations!
Anyway, if I'm not wrong, there is one important problem that you have mentioned before. How such LD spec can handle 3rd part activities (or modules, or, simply, resources) that aren't complaint with any of the LD supported specifications (IMS Content Packaging, IMS QTI, IMS Enterprise... and ADL
SCORM). For me, that's really a limited set of possibilities and I cannot imagine how other activities (and Moodle's activities are a good example) can fit into (be managed by) LD.
Being able to play (integrate) LD learning objects (not sure if it's the proper name) from LAMS, CopperCore... is really a nice possibility and I agree that it can be amazing to have all those stuff perfectly communicated with Moodle. My +1 vote for it, sure!
But I've some serious doubts about how LD could manage other activities that aren't the official activities supported by the specification. Does the specification support any extension to be able to work with 3rd part modules? Is there any possibility, under the specification, to define other type of elements. For example, the specifications supports SCORM content. Ok, I know SCORM is an important spec but, form my perspective, it's a 3rd part spec. And LD has included support for it. Is it an exception (an IMS spec using one non-IMS spec, that it's build from a lot of IMS-specs, I know

), or LD doesn't support to work with non-IMS activities at all.
In your previous message you said that "maybe we can model them using the spec". It seems that some work can be done within LD to map, model or whatever it was, to try to make the spec supports Moodle's activities (and vice-versa, some effort could be done too, in some areas of Moodle, to find the "nexus", without losing anything of our "Style of Learn" in the way).
Anyway, I think that step 1 (being able to integrate LD compliant products with Moodle) is a great objective per se. But step 2 (being able to integrate LD with Moodle's native activities) could be, simply, amazing.
I'm pretty sure that your next Portugal's meeting will be really interesting and I want to remark again that I believe both in Moodle and LD together (a bit more in Moodle, I must recognise it,

). Good luck and have a productive and innovative moot!
Ciao
PS: Perhaps I've said a lot of silly things here. Please excuse me because I'm an absolute LD's novice! I was trying to show my preliminary pros and doubts only.