Strange GFA outputs in uncompressed gfa #199
Description
The GFA seem to have cigar string operations longer than either node. Also have 0M edges where we couldn't find the overlap post-consensus. Consider also updating the gfa to trim back nodes to similar size pre-uncompression.