Michael,
It is very difficult to know how to remove aftershocks to get a true picture of background
seismicity especially when the catalog is as short as 100 or 200 years. Since seismicity
comes in cycles, you somehow have to describe where in the cycle you are.
For example, after the 1906 S.F. earthquake there was aftershock activity for about
10 years, then quiet conditions existed for about 60 years before activity resumed
and increased up to the Loma Prieta earthquake. So what your background is
depends on both when you take it and how long a period is involved in the
local seismic cycle.
When aftershocks are removed, most programs just leave a hole where the
aftershocks were or do not model the background seismicity well at all.
I worked with a colleague here for several years who collected (and developed)
more than 15 techniques for removing aftershocks from catalogs, did a statistical
analysis on their performance and found that none were successful in providing
what he considered a true picture of the background seismicity.