Those who are trying to filter are flooded with information that is redundant or they do not care about. But, countervailing views always seem to find a way in and most I know seek them out.

The majority of people only connect with a few sources and few friends. On many social networks the average number of connections is fewer than 20. When asking people why they do not have more they wish they could only connect on certain subjects.

This mapping starts getting at the reality at why and how people connect is real life. On line does not really replicate that.

When I said "intended outcome" I was more focused on the results/consequences of granularity. I understand what it is and why it's needed; I can't help but wonder, however, if "signal/noise as defined by the participant" prevents something--perhaps even the multi-disciplinary approach that lets us study topics like this.
That said, I wonder if the designer/developer has an right/obligation to define the experience according to their perception of participant needs--could my implementation of filtering options be biased by my perception? After reading Clay Shirky's recent talk on the surplus of societal potential (urltea.com/354n), I wonder if our ideas really look at both now and tomorrow.
Feel free to message me if you don't want to extend this thread. I watched this on your site which I found through brightkite, so feel free to find me there as well.