Add man page, numerous improvements

Redesign main loop to process all SAM alignments for each VCF call
Add hand-crafted test data
Add sort order checks
Factor out several functions to biolibc
Improve memory efficiency, now typically runs on 4-5 MB
Add MAPQ filtering with min MAPQ required on CLI
Collect stats on reads and alleles
Add script for clang profiling
Numerous other minor enhancements and bug fixes