This has been happening a long time. Print has been hurting for years. When I worked small market border stations, no local station did any local news whatsoever. In one market, the daily paper was pretty dominant but now struggles with 2 people in the building and 2 editions a week. The free weekly just went subscription because print costs have gone up so much. In another market, 3 of the 4 towns the station serves had their newspapers go under and the 4th went from weekly to 26 a year. The sole local station is barely hanging on and never replaced their news director when they got a new gig.
Indeed, and this has been something the journalism commentariat kept overlooking as long as regional prestige newspapers were raking in the money. Until they weren't. What was going on in broadcast, particularly radio, and in small-town or ethnic journalism, was beneath their attention. Those attitudes were also decades in the making.
It is going to be a harder landscape to figure out. The podcast that Kelly referenced used AI and the concentration of platforms as an example. When listening to that, my first thought went to the reader and the listener, many of whom are not well equipped in media literacy. This is where our educational system has been a big letdown. It isn't a new problem, and we could all slide by as long as there were trained and experienced gatekeepers to filter out junk "information". There are fewer gatekeepers now, and the ones remaining are less effective and some of their motivations are different and often more politicized and less bound to conventions of fairness that had prevailed for roughly the last 100 years. The Hardfork podcasters mentioned declines in information quality from Google search results; Google has no motivation to fix things because it profits regardless.
The other thing the podcasters mentioned was the imperative to provide a service that people want to pay for. This is where newspapers in particular fell down badly. They went in for long-form reporting tailored to the tastes of prize-awarding panels. That problem has mostly corrected itself, but in a way that has caused substantial damage to the viability of the publications that remain. It also doesn't mean pandering to an audience; that's a fuzzy boundary to be sure. Sometimes you have to tell people what they don't want to hear but you should have a good reason for doing so when you do it.
Regarding the Medill "news deserts" study, one detail caught my attention. It mentioned that only slightly more than half of public radio stations have local news coverage. This proportion is hard to nail down because it depends on whether you cont repeaters; leaving those aside, Medill counts "about 400" public radio operations, of which 213 provide local news coverage. It appears that Medill's definition of public radio means
radio stations that get CPB funding, which could exclude other outlets that provide some form of coverage, such as community-access stations. Oddly, Medill's map of public radio stations appears to leave out Iowa Public Radio altogether. IPR is sort-of-statewide - basically the eastern 2/3 of the state - and does provide local news coverage. I think, at least as far as this part of the study goes, more attention should have been paid to definitions and to data quality.