It's likely noone will ever see this, but it looks like what they need is a PageRank equivalent. Klout is AltaVista, counting numbers of followers and retweets. What they need is to be Google, distinguishing the true power of their connections. Google's key insight was in realizing that a link from the NYT was better than a link from geocities. Klout could do something similar, building a weighted social graph with the value of a given node determined not only by the number (and directions) of its connections, but also by their importance.
Ron Conway is more important than some "social media expert", because of who takes notice when he talks. The social media expert may have 100k followers and 5k retweets for every inane tweet, but Conway's utterances filter through the Internet's data channels, finding their way to other important people, TC, Forbes, etc. It should be relatively straightforward to judge the impact a given person has, and so make their Klout score more significant.
i had a friend who once observed that the more productive she was the lower her clout score -- i've since taken that to heart and see it as a measure of inverse productivity.
I have no scientific basis to back this up with, but here's my hypothesis:
The higher your Klout score, the lower quality of your connections/posts. Klout mostly cares about quantity, not quality.
The way Klout is going to make it's money is by identifying good evangelists. The best evangelists are not necessarily the best content creators (eg. Justin Bieber)
