My presentation yesterday contained a piece on integration of voice and presence into other behaviors/applications (for obvious reasons), and I ended it with a crude PowerPoint mock-up of what I thought the user experience might look like in future. I followed this with a screen shot of Gush to show the audience that this sort of thing is already taking shape often in the hands of dedicated developers using open source components. Telecom people typically don't seem to be exposed to these sorts of things, as far as I can judge, which is why something like Skype can come out of nowhere and bite them on the backside. Today a Triple Platinum value reader alerts me to another interesting example, Zimbra. This looks like it's dying to have some presence and voice capabilities added.
UPDATE: A mega-value reader points out that I failed to post a link to the mash-ups done by the company already as good examples of where they are heading. These are very impressive, I agree, but I guess I really meant more in the sense of something native to the client, and based on open standards.