. A great example that people like to use is that science did not think there was such a thing as meteors. People claimed to have seen them, but there was no proof. Then, proof was discovered and it became science.
A problem arises in the interim period between the first observations of a phenomenon and the moment when our official scientific institutions recognize the reality of the phenomenon. What about those unfortunate witnesses who observed meteors and were told that they couldn't have seen what they saw because the prevailing theory couldn't accommodate the data? They were witnessing an inarguably real facet of Nature, but because science hadn't yet processed that facet through the protocols of the scientific method, it was dismissed. And because in our culture science is the arbiter of truth, these witnesses had their truth reduced to the categorical status of error.
This gets especially touchy when the scientific establishment is for some reason resistant to change. Just because science
can accept new data and modify itself doesn't mean it always
does. Look at plate tectonics and continental drift. Alfred Wegener proposed the idea in 1912, but it wasn't widely accepted until the mid 1960's. That's almost half a century between the introduction of an idea and its acceptance into the mainstream. Look at Hugh Everett's many-worlds interpretation of quantum indeterminacy. Niels Bohr was the arbiter of value when it came to judging theories about quantum weirdness, and because he wasn't keen on having his reputation threatened by Everett, he wouldn't give Everett's ideas his blessing, and the many-worlds interpretation was relegated to the status of junk physics and pseudoscience. Now it's a viable option, but Everett ended up an alcoholic recluse because of his rejection by the scientific establishment (although he did enjoy a brief period of acceptance in his last years).
Of course, meteors and plate tectonics and many-worlds theory eventually made their way into mainstream acceptability, which confirms that in many cases science will eventually get around to correcting itself. But what about truths that, for whatever reason,
never manage to gain widespread acceptance, not because they fail to accurately describe reality but because the scientific establishment never gives them a fair chance (or even because we lack the technology to facilitate legitimate scientific analysis)? How many truths will be filed away as fictions and errors for hundreds, maybe thousands of years, because no one can get the grant money to study a subject that sounds too weird to be real? How many people will be told that they didn't see what they saw or experience what they experienced because it simply doesn't fit into the current, rigid scientific paradigm? In other words, at any given point in history, there are "fictions" and "falsehoods" and "superstitions" that could someday be revealed to be scientific facts, but there is no guarantee that this will
ever happen. Maybe a peer reviewer at an academic journal accidentally drinks decaf instead of regular one morning, and a paper that should be published gets rejected, and a potentially revolutionary line of inquiry gets stopped dead, never to be revived. Maybe a professor's pride gets in his or her way, and a brilliant young graduate student is dissuaded from pursuing a valuable line of research so that his or her dissertation advisor's life work doesn't get discredited. A thing can be true and real without having been given the stamp of approval by science, because science doesn't always exercise its responsibility to change itself when necessary.
Now, having said that, that doesn't mean that everything is true. Most "alternative" beliefs
are silly superstitions borne of ignorance and wishful thinking. But there may very well be something that we all laugh at, which science could, in theory, prove real, but for whatever reason(s), never will.
. 
