There's many technical, research pressures, and maybe political reasons, why the feeder weights are not balanced in such a way that at the end of the week, month or year, you'll be computing an uneven amount of time for each of the projects you've opted to participate in. One of several charts where you can see the overall present distribution is at https://bit.ly/WCGQLK1. Varying sub-project selections give substantially different distributions... for instance I've got FAHB+UGM on one machine, 8 cores, 1 day buffer. Not a single UGM is on there at this time [lower feeder weight due technical issues such as bandwidth limitations to the scientists in Brazil].