Data Packaging: Method to acknowledge bad readout_ids and missed timing

[kmharrington]

Right now, if an observation has bad readout IDs or missed timing, update_g3tsmurf_db throws an error every time it runs for 2 days straight. Most of the time these get checked pretty quickly and are acknowledged as "not for science." We need a way to tell the script to stop throwing errors on these observations.