The online activities may have been your first exposure to this kind of technology, although it is not exactly new. What I found out in my research is that it is mainly sound pedagogic models that are missing in order to make best use of the multimodal characteristics and the multitude of functions in this environment - otherwise live conferencing sessions may either become quite boring after a while, or they will more and more replicate didactic-style old-fashioned lecturing, ignoring most of the collaborative tools that are built into these systems.

By the way, I noticed that in the online activities, most Elluminate features were actually not used. I suspect this was because the environment was still new for the people involved. Sessions that take advantage of the full set of functions and that use these functions in a pedagogically sensible way do look quite different to what we did, in that they usually are much more interactive for each individual user. However, it's a bit hard to design such sessions, because - as mentioned earlier - good models or processes for mapping pedagogic frameworks to the full functionality of the web conferencing system are not really existing.

This is actually what we want to change in our research. Some background info where we are at the moment is available here:
http://www.lkl.ac.uk/research/mosaic/background

And the latest ALT newsletter has a case study from us:
http://newsletter.alt.ac.uk/e_article000783770.cfm?x=b8SDCQd,b3scdv19,w

I'd be interested in hearing your ideas and thoughts, and should you want to become involved in this research, there are certainly opportunities - perhaps a joint JISC-bid?

Tim