Ah, yes, important distinction (though I'd argue that what we know is less dependent on mainstream media nowdays). But the overall point, I think, still stands.
Unfortunately, we don't really know what we know or don't know in a lot of cases. We only know what we have evidence for, and we might have some insight into the quality of that evidence, especially if it is low quality.
One helpful approach is often to consider how many people have to cooperate in deception in the case that some claim about a deception taking place is true. This is pretty useful in cases like claims about the moon landings being hoaxes; people claiming hoax were, so far as I know, as a rule not involved in the space program when those landings took place. They're just random folks and so it is reasonable to dismiss them; there were a whole lot of people involved in the moon landings and they were very public-- and the Soviet Union would surely have claimed they were hoaxes if they had even the slightest indication that they were.
But this approach can also be misused: for example, the assertions about chemical attacks in Syria several years ago and the revisions of the OPCW reports about them and their divergent conclusions gave rise to controversies; the argument could be made that there would have to be a number of people conspiring to deceive the public in order for the final report to be falsified; we should expect a number of people involved in producing the reports to raise an alarm, otherwise it is a silly conspiracy theory. There is at least one crucial problem with that: a number of people involved in producing the reports did raise such an alarm. There were whistleblowers who were subsequently ignored and those who publicly raised concerns about this were subject less to counterargument and more to character assassination. This is one of the scenarios that you would expect to find if there were a deception; that doesn't mean there was one, but dismissals based on the difficulty of performing a deception fall flat. If there were a deception, there would be whistleblowers; there in fact were whistleblowers. In the case of a deception, other countries would have challenged the findings; other countries did challenge the findings. In the case of a deception, there should have been more whistleblowers; the people publishing the existing whistleblowers were subject to vile public defamation campaigns and marginalization, surely discouraging other potential whistleblowers from coming forward.
Another instructive example is the immediate aftermath of Operation Al-Aqsa Flood. Quite a lot of people would have to have cooperated with the deception in order for the claims about Hamas systematically using mass rape as a weapon of war to be a hoax, or the claims about beheading 40 babies or babies in ovens. These claims were published in Western media and repeated by American and other officials. The claims about babies were pretty easy to figure out as hoaxes: there weren't anywhere near that many babies in israel's casualty lists, and those that were had different explanations than beheading or baking: if you agree that knowledge is possible to have at all, it's all but impossible that these were not hoaxes. Of course, the existence of these obvious hoaxes might make other hoaxes easier to believe because of psychological priming and the fallacy of the golden mean. This one isn't as obviously false; maybe it is true?
The claim about Hamas raping lots of women is not as easy to positively disprove, and is still promoted in Western mainstream media as if there is clear evidence for it, and still believed by some because lots of people are a bit racist in their expectations about brown people; the idea that there are hordes of black or brown men eager to act on any opportunity to rape is a common white supremacist trope, and that trope still has power. They're coming for 'your' women (as is so often the case, and revealed at Sde Teiman, this is an inversion.) The New York Times has still not retracted Screams Without Words despite it being written by people with a combination of severe conflicts of interest and no subject matter expertise or relevant investigative experience, the evidence for its central claims (by ZAKA) being discredited in and by israeli publications and television, and no subsequent investigations confirming any of its claims. As recently as this month, articles in American media have treated disbelief about Hamas mass rape as per se antisemitic; skepticism of israeli claims during israeli perpetrated mass murders? It could only be Jew hate! (If anyone is wondering, I'm speaking of Olivia Reingold's article about New York City Mayor Mamdani's wife liking various israel-critical tweets, which is newsworthy for some reason.)
The proliferation of non-mainstream sources is both a help and a hindrance to having any idea what is actually true or false. The truth has an easier time getting out, but it also has an easier time being drowned in bullshit. It is good to be wary of all sources, and especially so of those that dominate your information environment. And don't forget that the establishment is perfectly capable of hiring additonal faces to read their desired narratives, wrapping it all in a lower production quality, and eschewing any recognizable establishment branding.