In article , Eric Flesch
writes:

The only part of FRW I have no answer to is the observed increase of
CMB temperature with look-back time -- that kills any static model
stone-cold dead -- except that a case for publication bias can be
made here, which I've discussed in the other thread. This includes
the question of how much our expectations bias our results, a- la
Millikan oil drop experiments. This is a broad unquantified topic
that normally one would want to avoid.

I don't think you've made a very convincing case here. As I pointed
out, publication bias can cut the other way too. Also, consider that
you know about the late publication of one paper. If you believe
publication bias is so rampant, then think of all the stuff you don't
know about. :-) Also, make sure that YOU don't have a bias here, i.e.
that you aren't giving too much weight to one paper and too little to
several others.