Improve flux measurments in the line list

It would be worthwhile to improve the flux measurements in the https://github.com/Subaru-PFS/drp_instdata/blob/master/data/lines/nist_all.txt  list.

This would improve our simulations and would help me so that I could estimate how close and strong nearby lines are. As Craig puts it in Slack `` It just seems wrong not to have reality-based line strengths, given what we know about the NIST lines.''