PyGRB workflow open points

[pannarale]

#4288 was merged but a few discussion points from this PR should be tracked. Namely

• new_snr_chisq in pycbc_pygrb_plot_chisq_veto returns the chi-square value needed to weight SNR into new SNR. Should it be moved to pycbc/events/coherent.py or pycbc/events/ranking.py eventually? It basically inverts get_new_snr from ranking.py with the caveat that the chi-square is set to a small non-zero number, when it should be zero, because pycbc_pygrb_plot_chisq_veto produces log-log plots so we are staying away from zeros. If anyone were to use new_snr_chisq in another context, this could be a problem, but I can always add a keyword argument to get around this.
• pp_files handling in pycbc_make_offline_grb_workflow. Quoting @spxiwh

I think it's a bad idea to assume anything about the order here, especially things like [-1] or [-2]. The FileLists were never intended to have a special order. That said, the case of a job that produces multiple output files was never supported well, and dealing with cases like this is ugly ..... Suggestions for improvement here are welcome!
However, I would suggest using one of the methods of the FileList class (https://pycbc.org/pycbc/latest/html/_modules/pycbc/workflow/core.html#FileList) to get these files. It's possible that a new method will be needed, as I think you'll want to search by "description", rather than "tag" (but I'm not sure). I think that file_list. categorize_by_attr('description') may produce what @titodalcanton suggests ... and would then be easier and clearer to work with.

• glsctables must be dropped from pygrb_postprocessing_utils.py as Completing pygrb xml to hdf5 transition #4419 is handled.

[pannarale]

The remaining issues are not about pycbc_make_offline_grb_workflow itself so they were moved to issue #4703.