expand chi_suppress_results to allow denom/num only, and/or allow different ranges

[rwbuie]

Currently chi_suppress_results does not allow user to follow our SOP for BRFSS data:

This suppression rule applies to the denominator and is for routine BRFSS analyses only. We also recommend suppressing estimates where the numerator is between 1 and 5. Given that DOH does not require any suppression for BRFSS data, APDE analysts may consider not suppressing results for custom analyses disaggregated by race/ethnicity or other demographics.