tiles-specstatus entries with incorrect EFFTIME_SPEC = 0 or NaN

[sbailey]

The following tiles in $DESI_SURVEYOPS/ops/tiles-specstatus.ecsv EFFTIME_SPEC=nan or 0, but were processed in Iron or Jura with non-NaN non-zero EFFTIME_SPEC:

TILEID LASTNIGHT  SURVEY PROGRAM EFFTIME_SPEC GOALTIME
int64    int64     str7    str6    float64    float64 
------ --------- ------- ------- ------------ --------
 41685  20220101    main  backup        145.9     60.0
 80724  20210328     sv1   other        131.7   1000.0
 81097  20210430     sv2    dark        247.1   1000.0
 81098  20210505     sv2    dark        180.3   1000.0

If they were re-processed in daily and tiles-specstatus.ecsv was updated, they would automatically be included in future prods. All of these were initially missing from Kibo. The first two (41685, 80724) were also in Iron/DR1, so we did the hand work to recover them in Kibo/DR2. The last two (81097, 81098) were in Fuji/EDR, but had already been dropped from Iron so we did not post-facto include them in Kibo.

Todo: reprocess all of these in daily and update their values in tiles-specstatus.ecsv so that they will be included in future prods automatically. The relevant nights/exposures are:

TILEID  NIGHT   EXPID 
int64   int64   int64 
------ -------- ------
 80724 20210328  82608
 81097 20210430  86750
 81098 20210505  87359
 41685 20211218 114606
 41685 20211220 114859
 41685 20220101 116294

Details, details: an initial attempt by @akremin on tile 80724 night 20210328 revealed that the processing table written at the time is incompatible with the current code, so it isn't quite as simple as "just rerun these tiles". It might be necessary to patch daily data with results from Kibo or Jura, and then update tiles-specstatus.ecsv.