That "official channels" bit has been the case, and a major failing of news organisations for over a century. Walter Lippmann and Charles Merz pointed this out in a 1920 New Republic article, "A Test of the News", commenting in large part on the New York Times's coverage (and failings) of the Russian Revolution. In particular was the issue of ideologically-tinged coverage, here anti-Communist principally. (Later the bias would run the other direction, particularly in the 1930s during the Holodomor.)
Lippmann and Merz note:
The analysis shows how seriously misled was the Times by its reliance upon the official purveyors of information. It indicates that statements of fact emanating from governments and the circles around governments as well as from the leaders of political movements cannot be taken as judgements of fact by an independent press. They indicate opinion, they are controlled by special purpose, and they are not trustworthy news. If, for example, the Russian Minister of War says that the armies of Russia were never stronger, that cannot be accepted by a newspaper as news that the armies of Russia are stronger than ever. The only news in the statement is that the Minister says they are stronger.
<https://archive.org/details/LippmannMerzATestoftheNews/page/...>
They continue to note the especially insidious nature of the anonymous source. The whole article is interesting reading, and bears strong parallels to events occurring today.
As do the practices Lippmann and criticise. Over a century later, news organisations still rely overwhelmingly on official (and unofficial) government spokespersons, and in the majority of cases treat such pronouncements as statements of fact, even where severe credibility issues are well known. Much of this is a result of availability heuristics (government mouthpieces are easy to find, and generally want to talk), reputation, and relationships established between journalists and sources. It is much more work to find truly independent, credible, and unmotivated witnesses.
And so, when things go pear-shaped, the official sources tend to become scarce.
Related to the disaster / comms failure dynamic I mentioned above: one rough proxy for determining how bad a widespread disaster in fact is is to look at where casualty reports have not yet been received. Again: capital and major cities typically preserve some communications capacity. Outlying regions are far more likely to be cut off, and by looking at the relative size and significance of locations that are making reports, as well as patterns of communications cessation, it's possible to make some inferences about total magnitude. Note that offical tallies of morbidity and mortality are based on received, credible, verified reports, which is to say, official statistics will almost always understate actual impacts, possibly for hours, days, or weeks, depending on overall severity. The 2004 Boxing Day Indian Ocean Earthquake and Tsunami comes to mind, with full official counts taking months or years to finally settle.
I'd long ago realised that this was a significant indicator of a crisis situation in many contexts. These range from military attack to natural disaster to political or business unrest.
The first indications of the atom bomb attack on Hiroshima were a) scattered reports of a "large explosion" and b) instant loss of all telegraph communications from a point some distance from the city centre inwards. Though the bomb struck early in day, it wasn't until that afternoon that an Imperial Japanese Army observation plane was able to fly over the city and surveil the damage, and the next day that the cause was known, after the United States informed Japan through diplomatic channels.
Similarly, when the HMS Sheffield was struck by Argentinian Exocet missiles during the Falklands War, the immediate effects were a loss of communications with the ship, and the first accurate information arrived, via heliocopter, along with the first casualties transported off the Sheffield.
In widespread natural disasters, particularly earthquakes, tsunamis, and hurricanes, there is often some communications from major urban centres, but even those are limited and often outlying regions are entirely cut off. I recall when following a major Chilean earthquake, the US immediately offered satellite telephones, which could be used to report on conditions from remote communities.
In business or personal relationship usually characterised by open channels, "no news is bad news" is a useful heuristic.
For start-ups and business, there's an almost ridiculously predictable progression of blog (and more recently: social media) updates, initially exuberant, enthusiastic, and often technical, shifting to highly-managed public relations releases focusing on business and social factors, to ... silence. The latter often ends with a "next stage in our story" post, i.e., "we're shutting down".
And in political and military situations, what used to be a fat channel of communications (though not necessarily useful or accurate) is cut off as chains of command become unclear, leadership and spokespersons scramble for safety, and rumour and gossip spew forth. That last is its own interesting mix: the genuinely confused or misinformed, often, but also those trying to influence or exploit circumstances.