While I support the PPM methodolgy over the diary method sample size remains an issue. Sure you are getting more days of listening from PPM than the diary but the pocket of participants is much smaller. It has an impact on the results especially if your P1s are not in the initial sample base. The other thing that continues to crop up is the variance in cume. There are still times when you see both the daily and weekly cume move significantly as much as 15 to 25% when there are no major external factors that influence audience to tune in or out. It doesn't happen all that often but enough to still be a concern.