Jon nails it! Find the best people in your domains of knowledge and subscribe to their RSS feeds (assuming you are not lucky enough to be on their team and work with them face to face and even then, their blogs will have great information that they simply can't covey face to face!). They will be your best source of real information. Second order sources such as PubSub and Feedster searches are great but a distant second in terms of quality of information.
The blog network is made of people. We are the nodes, actively filtering and retransmitting knowledge. Clearly this architecture can help manage the glut of information. More subtly, it can also help ensure that no vital inputs are suppressed because nobody has to rely on a single source. If one of the feeds I monitor doesn't react to some event in a given domain, another probably will. When they all react, I know it was an especially important event.
The resemblance of this model to the summing of activation potentials in a neural system is more than superficial. Nature knows best.