cobraId 2182, fiberId 820 is consistently missing from cobra_match table

As reported by rhl, in our calibration set from July, one fiber(fiberId:820, cobraId 2182) that was supposed to be homed is labelled as BLACKSPOT in the pfsConfig which shouldn't happen.

When we take all calibration we move all cobras to home position and then take a mcs exposure. The expectation is to have a match for every sane cobra, 2356 in total.
I checked pfsConfig from April and it was the case, but in July I've found that cobraId 2182 has consistently no match, we moved to home several times.

Either something broke for that particular cobra and there is no flux in the MCS image, which will interest yuki.moritani very much, or there is an issue in the matching code.

karr chyan can you check in the MCS image ? here are some visits, all taken at HOME :

/data/raw/2023-07-20/mcs/PFSC09691500.fits
/data/raw/2023-07-20/mcs/PFSC09700100.fits
/data/raw/2023-07-21/mcs/PFSC09703500.fits
/data/raw/2023-07-21/mcs/PFSC09707100.fits
/data/raw/2023-07-21/mcs/PFSC09716600.fits
/data/raw/2023-07-22/mcs/PFSC09738000.fits
/data/raw/2023-07-23/mcs/PFSC09759600.fits
/data/raw/2023-07-23/mcs/PFSC09759800.fits
/data/raw/2023-07-24/mcs/PFSC09782900.fits
/data/raw/2023-07-25/mcs/PFSC09806100.fits
/data/raw/2023-07-26/mcs/PFSC09843200.fits
/data/raw/2023-07-27/mcs/PFSC09882000.fits
/data/raw/2023-07-28/mcs/PFSC09907000.fits