Insufficient Calibrations Found; arc darks

[Nat1405]

This bug was found when trying to reduce GN-2006B-Q-107. Arc darks for 20060102 don't exist for that particular program (or any program), but they do exist for a different program on the day before and the day after. Nifty should probably be able to look for missing calibrations in all nifs programs one day before and after the observation in question. After that it should error out and let a user try to find appropriate calibrations.
Essentially, the pipeline should:

• Query a data source for $calibration_types, matching $observation_day, $type=dayCal where $observation_day is the original observation date +- 1 day, and
• Download this data into the appropriate rawData directory
• .... Restart the sort? Hopefully sort is fully idempotent.