In article , Eric Flesch
writes:

The general expectation is that the CMB temperature INCREASES with
redshift. How many papers find this, and how many find something else?

Sure, Phil, but consider these points from this incident:
(1) Team A (1999) found disagreement but did not publish
(2) Team B (2012) found agreement and published.
(3) Team A published in 2012 to balance Team B's finding.

Of course, you are speculating on the motivation in (3).

Therefore, how many other such findings have gone unchallenged simply
because there was no other "Team"? Note also from this incident that
papers are more likely to be published when they agree with the
current wisdom -- Team A did not publish for 12 years.

Again, you are making an assumption without much justification then
extrapolating from it.

So it is
entirely plausible that efforts on this front go 50-50, but the
affirmative ones publish and the negative ones do not.

I agree that publication bias is a serious issue, but it cuts both ways.
A&A has (or at least had at one time; I don't know if it is still true)
a policy of not publishing "boring" results, i.e. those that agreed with
current wisdom and offered nothing new. In some cases, NATURE tends to
be less critical of surprising results and in some of those cases
wouldn't have published them had the result been more in line with
expectations.

Not really relevant here, but IIRC, 1830-211 is a gravitational-lens
system. Any static model needs to QUANTITATIVELY explain the huge
amount of gravitational-lens data.

So you mean, we need to explain the lensing angles in the absence of
the modelled "dark matter". Good point!

Not just that. The basic idea of gravitational lensing is quite old,
even older than relativistic cosmology, but the theory was put in place
only after relativistic cosmology was established. So, it was developed
with relativistic cosmology in mind, and seems to work quite well. The
analysis of a gravitational-lens survey is a huge undertaking, and the
fact that the cosmological parameters derived from it agree with those
from much simpler (at least conceptually) tests is a huge argument in
favour of the underlying assumptions being correct. Any alternative
model would have to explain why it works if the underlying model is
different. Another example, Arp and others have questioned the
cosmological nature of some high-redshift objects, particularly QSOs.
But there are many examples of lensed QSOs, and in all cases the
redshift of the QSO is higher than that of the lense. If the high
redshifts had another origin, we would expect at least a few examples of
a source with a lower redshift than the lens. (In some cases, Arp has
attempted to explain gravitational-lens systems by claiming that the
multiple images are multiple objects emitted from the lensing galaxy, in
line with his claim that such ejected objects have large
non-cosmological redshifts, but this seems like a HUGE epicycle to save
the appearances of his ideas.)