Update extra-pixel correction logic in PfsMapper:std_raw()

Earlier versions of the CCD DAQs generate an extra pixel at the start of each amp's readout. This shifts all subsequent pixels from that amp over by one. It manifests as a shifted column, but is just a shifted pixel.

The std_raw routine optionally applies a correction for that effect, based on the existence or value of a now obsolete header card.

For recent data, this correction is being applied when it shouldn't be.

Retrace the history of the issue and rephrase the logic in PfsMapper.py to apply the correction only when it should be.