The performance of arfgen depends mainly on the number of points in the detector map that are enclosed within the said regions. This means that the coarser the detector map, the less CAL calls are made, and the less accurate the resultant ARF is. At the other extreme, a detector map that is too finely binned would result in a large number of CAL calls and would affect the execution time of the task. Ideally the binning should be such that it reflects the spatial scaling of instrumental variations across the detector. The current version of this task does not have any feature that varies at scales smaller than 1 arcmin, so the user is recommended to use detector maps with bin sizes around this value. Once pile-up effects are accounted for by this task, the bin size for the detector map would have to be much smaller in order to oversample the PSF ( arcseconds). It is expected in such cases the user would mostly be interested in a single point source, so the total area enclosed by all the regions, and thus the number of points in the (filtered) detector map, should be relatively low.
The execution time is also dependent upon the actual size of the source region if chip gap and bad pixel corrections are switched on (withbadpixcorr=true). In fact this becomes the dominant contribution for large source areas.