I included the coating thermal noise (CTN) in the estimate made in entry 25039, and find that this changes little the conclusion.  It seems that the unidentified noise source has a slope close to 1/f.  Increasing the CTN estimate by 1.4 changes the conclusion, in that noise is more like 1/f^2.  I do NOT mean to suggest that this indicates higher CTN; that extra junk could be anything.  What this does mean is that we cannot use H1 to set a better limit than this (yet).