Give a way to report width of xcorr peak

[eteq]

At @clidman's suggestion I'm bringing this idea up here.

We're using marz as a redshift/quality-checking tool (fantastic for that, by the way!), but one of the things we care about is at least a ballpark uncertainty estimate on the redshift. A straightforward way to do that would be to use the width of the xcorr peak of the reported redshift (say, using a flux-weighted std estimate centered on the peak ), and then include that in the output file. Does that seem reasonable/possible?