Yes, the Internet is a really big deal, a fact which Google trends makes very clear to me. But big brands, as Yglesias notes, are doing really well. Magazines have done alright too. It's the combination of the Internet plus all sorts of other stuff that are putting journalists out of business. After all, virtually every newspaper now puts their content on the web. They've realized that they're not in the business of selling paper, but rather of selling media content, which can happen in any medium.
Online media distribution has very different economics though. For one, you don't have the crazy diseconomies of scale which have supported cozy regional oligopolies. If I'm on the web and searching for news, I might as well go to the best source of news, while in my hometown I might be limited to a few local papers (no, you don't want that mail subscription to the NYT). The best online sources of content have generally done pretty well with respect to page views. The strategy of many news teams, however, has been to respond to budget crises by cutting back writers, which pushes more people to other online sources, which hurts the budget more, which causes another death spiral of cuts... the WSJ and NYT, however, have kept up pretty good news teams and seen plenty of visits their way.
But newspapers would still be fine if they were able to just move their customers online and still make similar ad revenue per customer. They'd perhaps be better off since the cost of a newspaper is pretty close to the production cost. The big problem is that Internet advertising is nowhere near as profitable as print advertising. You behave very differently on the Internet than other media places, and advertisers don't really know how to deal with that. They might fix this one day; the first TV ads were rip-offs of radio ads and no one paid them any attention. Perhaps the defining characteristic of the Internet is the network structure of links and the quick diffusion between sites; it's not crazy to think that someday we'll want to click on those links, since we click on everything else we see without thinking about it too much. Probably the best way to get there is to data-mine personal information to deliver links that look like things I click on all the time. I click on the news stories at the top of my Gmail account, for instance. People are going to lose a lot of privacy, hands will be wrung and the occasional sob story will hit Newsweek and people will be better off. But right now Internet advertising is less profitable for newspapers by a substantial margin; it's not that they've lost customers so much as they've lost the profit/customer.
The Internet also takes away a lot of the reason for their being a newspaper, which is after all a bundled product of various smorgasbord news items. This used to be the most efficient way to deliver a mass of information to a person. These days, people tend to use specialized sources--ESPN (or, even better, deadspin) for sports, gawker for entertainment, Slate for contrariness, blogs for opinion, and so on. There are a handful of aggregators, and of course the best media brands with international coverage draw attention, but there's much less reason to use a mid-size town newspaper for this purpose. There's maybe less serendipity of coming into wayward news events online, but really people use the internet for a lot of random stuff. If historical evolution went from the internet to the newspaper, old hands would complain that young folk no longer get the educative experience of browsing Wikipedia.
Some newspapers have responded by increasing specialized coverage to appeal to wider audiences, and letting reporters write blogs to create some personal information channels. In the long-run, the Internet is a very un-democratic institution which makes everything follow a power law. So a few blogs dominate, and many of these tend to be run out of big media institutions (and staffed by Ivy Leage grads...); the remaining big blogs are mostly legacies. Again, the problem lies in monetizing all of these people. It tends to be only possible with extremely rich people, which is probably why Murdoch decided to keep the WSJ behind a wall (while silently redesigning the site). Yeah, that leaves you without ammo in the daily link barrage, but your proprietary blogs can still participate in that while you focus on profits.
This has all been great for consumers. Following the general trend of globalization--kill local options while increasing choices for everybody through trade--it's easier to get quality everything than every before. Probably investigative reporting will take a big hit, since bloggers are better at shifting through things in the public domain rather than adding original reporting, but who really cares. The top-quality newspapers will do very well, as will niche publications and anything that caters to rich people.
The big unresolved problem in my mind is coming up with decent personalized aggregators. Sites which take a person's given preferences to deliver things that 1) All their friends are reading 2) Appeal on the basis of 'you really should know this' 3) Provide some random flavor 4) Are similar to things you have read 5) Go beyond fetching stories to summarize ongoing debates 6) Provide "Internet criticism;" broad, meta-criticism of overarching trends and themes. Of course, you'll be able to make this as personalized as desired, vote up/down on stories, send to friends, etc. Google News doesn't really cut it--humans might need to get involved at some point. Naked Capitalism is nice for some finance/economics stuff. Newser is pretty interesting too, but there's a lot of room to get better. Slate does some of this, and a service similar to Slate, but for everyone and user generated, recently went bankrupt.
Basically I'm waiting for the point where I berate people for not reading/creating blogs like normal people and instead twittering or some nonsense.