Why is accepting a scientific consensus rational?

Because at one time the scientific community subscribed to the steady-state model of the universe (for which they also had reasons/evidences) but now that has been completely overhauled and replaced with the Big Bang model. It seems to me that scientific consensus is not always immune to error (even if scientists furnish evidences because later on they could discover counter evidences, which compel them to rethink their long-held views). Before Edwin Hubble discovered the existence of other galaxies, it was widely accepted that ours had been the only one. Einstein revolutionized physics. Prior to him we thought the universe was only governed by Newtonian mechanics. Furthermore, if scientific consensus is only sometimes correct, how does the layperson decide which consensus is and which isn't? In fact, how does anyone? There doesn't really seem to be any certainty involved.

The answer to your question has to do with the difference between being right for the wrong reason and being wrong for the right reason . To state this another way, even though there may be excellent reasons to hold a belief, the belief itself may still be false. Science is NOT about certainty; it is about probability. Likewise, inductive reasoning is based on probabilities, not certainties, so when there is an overwhelming scientific consensus by experts on any given subject this means that the consensus is likely to be correct given the totality of the available information. Granted, there are a few salient examples where the dominant theory was replaced (e.g., steady-state model of the universe), but this is dwarfed by the number of theories that are simply improved upon and made more accurate with time (e.g., Newtonian mechanics is still valid and only improved upon Einstein's theories and the discovery of other galaxies did not change the fact that our galaxy still existed—just our assumption that ours was the only one).

One more important point—these days, legitimate scientific dissension is found in published, peer-reviewed journals and academic conferences. Unfortunately, public opinion is swayed far more by those who make YouTube videos, post blogs, make Facebook memes, set up their own "institutions" or "foundations," and appear on certain news programs—even if their ideas have been thoroughly refuted already within the scientific community.

For those of you who are interested in a different but very interesting perspective, check out Stephen Hawking's model dependent realism that focuses on the usefulness of a model or theory rather than what we call the "reality" of it.