True, but as pretty much all scientists exist within the same paradigm repeatability of results helps a little but still gets us nowhere near objectivity.
It doesn't? What does? How else would we determine objective facts.
Why do you think repeatability only "helps a little". If there are those who disagree with the results, that is, those whose views would bias against the results, and they repeat the results doesn't that do more than "help a little"? Even if they replicators are operating with the same overall views they aren't just going through the replication for fun. They are trying to see if it does work.
You'd have to show the particular paradigm and how it presents a problem to a particular result to be able to demonstrate that there is a real problem.
The danger of individual bias is, of course, recognized. At any time when there is room for experimentor bias the experimental arrangement should be double blinded. This allows for results to be evaluated
without an awareness of how that affects results.
In other cases single blinding is adequate.
Could you explain how the current process only "helps a little" and offer precise examples of the difficulty?