What you say is technically correct for a very narrow span of time, but also one of the most pernicious myths about the finances of the gaming industry.
The article you link is from when the 360 first went on sale in 2005. The 360 remained MS's "main" console until late 2013. Production costs fall wildly over that time. Indeed, in the traditional MS/Sony model of selling consoles, you sell at a loss for about the first 12-18 months, then as unit cost reductions and economies of scale start to work in your favour, you keep the console selling at a more or less neutral level for the rest of its life-span, reducing the retail price as costs fall further.
Where do they make the money from? Xbox Live subscriptions, first party games etc are a small part of it, but only a small part. Most of the money - and it is a lot of money - comes from third party game fees.
See, when you buy a console game as "new" (rather than pre-owned), a large chunk of the sale price goes directly to Microsoft, Sony or Nintendo. On a full-priced game, this tends to be in the $10-15 range. Historically, this has explained the price differential between console and PC games - though with Valve now taking a similar cut of most PC game sales, who knows how long that will last.
The platform owner has spent next to nothing on those third party games; in most cases, it only gets involved at the certification stage. So it is, for the most part, "free money". And with series like Call of Duty, FIFA, Madden etc racking up the sales they do, it is a lot of free money.
So the trick is attracting third parties to the console. To do this, you need to have either a large current installed base, or the promise of a large installed base to come. This is why console manufacturers are happy to sell at a loss for the first year and often to take a loss (or at least a risk) on funding first party or platform-exclusive third party games - the Halos, Gears of War, Killzones and Gran Turismos of the world. Those are the bait to lure in the early adopters to get the installed base growing to get the third party developers on board.
The other business model is the one that was previously (but not currently) used by Nintendo. In the SNES, N64, Gamecube and Wii generations, as well as with its handhelds up to and including the DS, Nintendo sold platforms at a profit from day 1 and focussed much more on first party games development. This actually worked pretty well for a long time; they made megabucks on the SNES (which also had a lot of third party support, so win-win there) and even when the Gamecube ended up with poor sales, they were still able to turn a profit on it.
But around 5 years ago, this model started to break. The Wii was essentially dead by 2010; console sales were slowing to a trickle (after a few phenomenal years) and despite the huge installed base, most Wii owners (a different demographic to that on other platforms) did not buy many games, so third party developers abandoned it. Then came the 3DS launch.
The 3DS is doing ok now. Well in Japan, so-so in the US and Europe. It's on course to be a kind of PSP-level success, which is ok (the PSP actually did much better than is generally realised, largely on the strength of Japan). But the 3DS's launch was actually a bit of a disaster. For months after launch, the damned thing just wouldn't sell - and price was a big part of it. So Nintendo reversed historic policy and slashed the price; for the first time in its history, selling console hardware at a loss. It didn't remain at a loss for long; only 6 months or so until it got onto a neutral footing - but it was enough to bury Nintendo's historic strategy. Console sales improved, third parties moved in (particularly Japanese developers, many of who shy away from the high cost of developing for home consoles) and Nintendo's losses (the first in the company's history) were reduced. When the Wii-U was launched, it was launched with a traditional Sony/MS style pricing strategy; sold at a loss at first, before moving to neutral pricing after a year or so. In the Wii-U's case, for a variety of reasons, that failed to get it a good installed base and Nintendo now has an outright disaster on its hands - the hardware isn't profitable, third parties have left and the only business left is selling first party games to a relatively small user base.
The fun thing about this cycle is that following the poor launch of the 3DS and the disaster-launches of the Vita and Wii-U, some developers bet against the PS4 and Xbox One succeeding. 2k, in particular, committed itself to a strategy of ignoring the new consoles, while focussing development on 360, PS3 and PC. That's cost them a lot of money, with their sales significantly down on a few years ago. Borderland: The Pre-Sequel in particular has been a bit of a sales disaster, with a belated port to the new consoles jsut announced.
What 2k (and others) forgot is that installed base is important, but so is the propensity of that installed base to buy games; and early adopters of new hardware tend to buy a lot of games.
But yeah, in the big picture, installed base is critical and the fact that console manufacturer's take a loss for the first 12 months or so isn't particularly relevant.