Given the same data and consistent application of the scientific method, it should be expected that a field will converge on a “consensus,” whether of the form “hypothesis Y is probably true” or “hypothesis X and Y might be true, but Z is definitely not” or “nobody knows anything beyond fact Z.” Since this is what is expected to happen, when a field can’t reach a consensus, observers will give less credence to the claims of the field, thus, for example, when so-called “fact-checkers” give different ratings to different statements, the credibility of all is undermined. Thus, fields have an incentive to converge on a public consensus even when they privately disagree; once a consensus is reached groupthink and punishment of dissidents will prop it up. This gives people outside the field a reason to distrust it.
Our descendants may not have this problem, if they are all individually so intelligent that they needn’t put their trust in some academic caste. However, they might just use this intelligence to discover even more difficult puzzles, puzzles we can scarcely imagine, just as our ancestors could not have imagined quantum mechanics. In addition, many fields could require our descendants to digest decades worth of information before they can really understand them, decades most would rather not spend. Thus, academic experts may still exist.
However, if spread through the galaxy, their ability to reach a consensus will be undermined due to the vast communication delays. Expanding humans can bring existing consensuses with them, but if two civilizations a hundred light years away independently find some new data which must somehow fit into the consensus theory, they will be unable to collude on a consensus view. Thus, it will be a natural experiment in whether they are really applying the scientific method. The same would be the case in the unlikely event that humans encounter intelligent aliens, who would almost certainly have some idea corresponding to the scientific method.
Our descendants may want to purposely cause this situation, to prevent the possibility of self-deception and increase the trust in their field. For instance, if a human colony finds a primitive extraterrestrial civilization and wants the best theory of how their languages are related, they may decide to demur for a time the broadcasting of their own theory(which would be a natural starting-point for the consensus viewpoint) and instead only broadcast the raw linguistic data, waiting to see what the surrounding human civilizations have to say about it.