I just saw someone comment that "Twitter would be improved by better list functionality". For years my only experience of Twitter has been through a list of about a hundred people who I trust to add more signal than noise. I have also been adjusting my Facebook lists, after watching Robert Scoble's enthusiasm for adopting that technique, to raise the level of stories I see and reduce the number of pictures of cats.
As soon as we get "improved" information systems we are exposed to too much information, feel swamped, and feel the need to improve signal and reduce noise. We are then faced with the problem of classification. Do I classify someone as a person whose views I value on technology, a friend, someone who spots quirky stories, or all of the above? If I put them in a technology news list that I improve by "liking" technology related posts, what do I do when they then post a heart warming human interest story that I also want to like? The people I follow aren't one dimensional. The appeal of social tools is that we get to see more than just one aspect of the people we connect to.
It's the same old issue that was meant to be sorted by portals and executive dashboards. As soon as I think I have reduced noise in my system I begin to worry that I am missing out.
The solution is partly to use different tools for different purposes. This is why I prefer an ecology of diverse tools in a business environment rather than the all encompassing enterprise social tools that are the current fashion. It's a bit like "real life". We expect to meet different people in different physical contexts and manage our expectations of different conversations in this way.
The problem at the moment is that in the competition between Twitter, Facebook, Linkedin and the others for our time and attention, they are all beginning to look and feel the same!