This is the table of statistical requirements for various analyses.  Later I will document here where we came up with the particular numbers.

Note that one table is number of events required, and the other is, assuming George's estimate of events/10-hour-day, the number of weeks it would take, to get a meaningful measurement of that observable.

As noted, these numbers are not meant to be very "inflated for safety"

there is a .pdf file and an excel spreadsheet (.xls file)