I run distributed computing projects because it's an easy form of charity and a means to make a contribution to the advancement of science without actually doing any benchwork. I only have one or two computers to do crunching, and so I choose just a handful of projects to work on so I can see my graphs go up in a reasonable timeframe. :-) I try to keep a couple of big well funded projects so that I always have WUs, but stock my stable with little "underdog" projects so I can have fun. I try to have representatives from both life sciences and physical sciences.