Convergence calculation hangs when NS/BH populations are run with cosmic-pop

[katiebreivik]

If a user specifies that they want to simulate NS and BH populations in a single cosmic-pop run, the Knuth binwidth calculation takes a long time (sometimes days (!)) to compute the binwidth for the mass histograms in the convergence calculation. This makes the calculation intractable and we should probably generate some warnings for folks who are trying to do this.

Some options for warnings:
(1) warn when 13 and 14 are specified for final kstars
(2) warn when mass ranges below 3Msun and above 3Msun in convergence calculation

An option for future catches:
(1) split the match calculation for NS/BH masses up for > 3 Msun and < 3 Msun similar to eccentricity catches