TFA is, I'm sorry to say, complete drivel. It ignores two key considerations.
First, Valve's platforms - Steam-on-PC/Mac and the forthcoming Steambox console - are home platforms. Where the pay-to-win model has achieved some success (and even there, the successes are outweighed 100-to-1 by the failures) is on the mobile platforms, where people play for snatches of a few minutes here and there. PC and home-console gaming remains dominated by more substantial offerings, with more significant development budgets and (frankly) a more discerning audience.
And the second point is just that; games cost money to develop. Quite a lot of money, these days. We're already seeing an increase in the RRP for games on the new consoles, which, irritating though it is on one level, is probably something the industry has needed to do for a while now. Long story short - nobody is going to be rushing to give these games away for free. If Valve wants a console, retailing at a per-unit profit, whose selling point is a mass of free titles (and I don't believe for a second that it does) then it will need to throw a massive, unprecedented subsidy at game developers. And that's just not going to happen. We've seen what happens when you try to launch a console whose selling point is the kind of games you actually can give away for free or near-free. It's called the Ouya.
Which, as we all know, is doing just splendidly. Or not.
What Valve's move does unlock the possibility of is smarter and more responsive pricing for games. And this is where there's real potential for the industry to do better.
Historically, we've sold games as though they were movies. There's basically one price point when they're new and another for when they get a budget re-release. Ok, indies and the like have always played around outside that system, but the actually relevant commercial developers have had very fixed price structures. What Steam has moved towards - and seems set to move further towards - is pricing that can price games more accurately reflecting the value they offer, their review scores and their week 1 sales.
Bricks and mortar retail stores sometimes try this, but the way in which they purchase stock and are insured on those purchases makes it a last resort for them. The ability to flex prices rapidly at the publisher level is much more useful. If you have an Elder Scrolls style RPG with a huge development budget and hundreds of hours of game-time, then go at $80. If you have an average sized shooter, perhaps in the $60-70 range. If you have a 2d platformer or sh'mup, then perhaps you should be thinking more about $20-30 for your first release.
Nintendo, in particular, desperately need to learn this lesson. My theory on the unnoticed reason behind the Wii-U's continuing disaster is that it's just too obvious that Nintendo's pricing is vastly out of whack with the value their games offers. Ok, the $60 price-point might be ok for something like Super Mario 3d World, but is it really appropriate for 2d platfomers (Donkey Kong, New Super Mario) or HD remakes which sell for $30 on other platforms (Zelda: Wind Waker).
No long slashdot post would be complete without a car analogy, so I'll say that game pricing needs to be less like movie pricing and more like car pricing. It should have a much wider range and be more responsive to features like production costs, quality, features, brand and image.