Stephen Abrahm's Reference in Transition examines some common (and less common) reference scenarios: "traditional" aka status quo, information commons, learning commons, embedded librarians, partners in action, the remote librarian, team players, the retail librarian, shoulder to shoulder (teaching & training), avatar based/virtual worlds, virtual librarian (no onsite f2f presence), emergency/on-demand librarian, aggregation of user experience, aka crowdsourcing, and finally, the mandatory -- all of these.
I think for the VL/avatar based librarian to become a viable choice, those platforms are going to have get easier to use. SecondLife is very computer resource intensive, and it's not easy to learn to navigate the SL world. Smallworlds seems like it may be a viable option (flash based with no software to download), but really can anything take down SL at this point?
I definitely have to disagree with this statement, though:
For instance, the OPAC and ILS systems don’t suck for library workers. They were built to meet our specific needs — library management, transaction processing, inventory systems, etc. When we moved an internally oriented tool out of the backroom to make it accessible to the “public,” we did a good thing. The unintended consequence of public OPACs, however, has been to teach us that end users have different needs and processes for discovery and navigation than library workers — especially in the virtual digital world.
I haven't met an ILS (and its OPAC) that didn't "suck" in some way. None of them are perfect -- as anyone who has actually done work in one knows -- as anyone who has done a search knows. Just because you might know more about how the ILS works, doesn't always ensure success nor a good experience. Ugly? Poorly designed? Lack of functionality? You betcha. I usually cringe when I go poking around in a library's OPAC. I also don't think our needs are so different from library users -- because after all, we are library users, too.