I don't think we can necessarily say this is the case.
I can imagine the ability to create new bias through repetitive propaganda spam reenforced with real time trolls pushing the same bullshit over and over, then attacking them for any weakness in their stance creating a second guess to the actual facts. Then over a few more years those bubbles start to harden and they get trained all the time to combat rationality and non-populous messaging.
This is true, we can change, but it is also true we tend to believe the first thing were hear and then discount subsequent information that contradicts this. Just as biases can be confirmed, they can also be eroded, this is both a problem and a solution.
I have no idea what this means. Science is replicable. I think you may be confusing bullshit youtube videos for science. To actually understand the science means you had to have earned that knowledge, it doesn't just happen by osmosis.
The scientific process itself is a process of removing biases, facts not beliefs. The standards of evidence differ for different branches of science, reproducibility is empirical in astronomy and double blinded placebo studies are not required in most branches of science to determine the facts. In areas where results are not easily ascertained or judged, the rules have to be tighter. Reproducibly is one thing, causation is another, one can reproduce results without understanding causes and causes are what science is all about.
In physics it simple, the data must agree with the theory or the theory is wrong. When new data becomes available that contradicts an existing theory, then that theory is wrong, though sometimes usable until something better comes along. Newton's theory of gravitation is wrong, but his calculus based on that theory is still used, it is good enough for the purposes it is used for. Like wise relativity breaks down on smaller size scales and quantum theory must be used. All three of the traditional theories in physics are technically wrong, but correctly predict things and are thus useful to engineers. Newton, special relativity and quantum theory are all wrong, but useable and to an extent correct. They offer a peek behind the curtain to the true nature of reality.