optimize for usage in metapipeline

[sorelfitzgibbon]

          In order to optimize this pipeline, it would be helpful to start to specify what QC results would trigger a cessation of the metapipeline run.   For example coverage calculations take the longest time and may not be necessary before the gatekeeping.  Results from `SAMtools stats` should be sufficient, e.g. read QC and count, mapping QC & % duplicated.

In this case, perhaps this pipeline could be run twice within metapipeline, first targeting the gatekeeping information (SAMtools stats) and second filling in the rest (FastQC and coverage (mosdepth or collectWgsMetrics)).

Originally posted by @sorelfitzgibbon in #62 (comment)