[daip] SPLAT, SPLIT reference frequency problem

[Eric Greisen]

I have revised SPLIT and SPLAT to get the average frequency in all the
channels/IFs being summed to gether.  I also took the AVIF option out
of AVSPC since it claims to be doing multiple FQIDs and I have no idea
how to fix that.

I added comments on the fact that weighting and editing will make the
effective frequency at each sample not exactly the frequency used for
this correction which is an unweighted and unflagged average.
Astrometry should probably think about that issue.

Eric Greisen