When we update the latest snapshot after a snapshot task has run, we blindly update the distribution.   Instead we could fetch the distribution and compare the publication to what is expected.   If it matches, no pulp task is needed.

This is important as all distribution update tasks are processed serially.