So let's see how much of this I can remember.
The session opened with a discussion of color theory as applied to maps. Not really anything that we don't already have in our resources section. We had a report on Web3D standards, and how a set of geospatial specifications is being written to allow GIS systems to procedurally draw 3D graphics in a web browser. More on that in a bit.
Then a researcher presented the results of an experiment to permit different sorts of map servers using significantly different presentation paradigms to communicate with one another. In essence, they developed a system where you can use data from, for instance, OpenStreetMap in Google Earth. I believe he was using four different server types and four different clients. They did not manage to make every connection, but I think I recall that they got Google Earth to read data from all four servers. That experiment was called 3DPIE.
Another researcher presented a piece of software called bitmanagement, which sounded like it did something similar to RR's Viewingdale, except that the maps are presented in a web browser, and again they used a 3D rendering system, although it didn't sound like it was the W3D standard.
I also attended a session about simulation with computer graphics, in which a scientist from ESRI demonstrated CityEngine and some of its planning tools, as well as giving a demo of its Level-of-Detail culling functions, allowing the user to wander a 10 mile x 10 mile 3D environment, rendered in real-time with Web3D. No client necessary, only an HTML 5-compliant browser.
Incidentally, he mentioned that ESRI is hiring "content producers." He was a little vague about what that meant, but I imagine that any GIS-skilled cartographers among us who'd like to work in the field might want to give them a call. He was from their Zurich R&D office, but they have branch offices all over the world.