Hmmm, well I tend to disagree with a lot of what is said here. The biggest area is probably the idea that a game explaining itself and the plotline/storyline is a bad thing, and can't ever work out. The problem is that a lot of game developers, or heck developers of media in general, use the whole "we want to be mysterious" thing as a crutch for bad writing and development. Pointing towards how things might lead towards some overarching, sinister truth, while tossing in logical inconsistincies towards people who try and piece it together is a heck of a lot easier than actually developing a solid storyline and then telling it in a mysterious fashion that ultimatly leads to a fulfilling conclusion. The problem is that your dealing with what are basically hack writers shooting themselves in the foot trying to wrap up their work in a sequel. The problem is that almost every video game series that tries to be mysterious winds up being "Lost", oftentimes trying to buy it's way out of the problem by being increasingly more obtuse until they wind up at a point where they can't do that anymore, the series has to come to an end, and they are left facing a mess of their own creation.
To put things into perspective, it's possible to have creepy and mysterious atmosphere that seems to make no sense, but then wraps up in a satisfying conclusion that answers everyone's questions and leaves people thinking "OMG, that was brilliant" rather than wallowing in some kind of odd psychobabble, or revealing the protaganist was crazy all along, or relying on surreal time travel manipulations and "WTF" reveals along with ignoring most of what had been established up until that point. The thing is that we've seen mystery and horror writers do it right on a number of occasions. Agatha Cristie, Sir Arthur Conan Doyle, Steven King, Dean Koontz, Clive Barker, Michael Slade and others have all become pretty famous and successful by managing to pull this kind of thing off, even though they have had failures along with their successes. A lot of their best work might seen quaint in retrospect, but only because we all know the twist and how it all falls together.
If I had to point a finger at the real masters of this I'd actually give a not to Carolyn Keene, and Frank Dixon (which is actual a pseudonym of a lot of differant writers, similar to how Michael Slade is actually more than one guy). Carolyn Keene and Frak Dixon did "Nancy Drew" and "The Hardy Boys" respectively. I give them the nod because while they write children's literature and young adult fiction, they are masters at taking wierd stuff and impossible events, and then tying it all together at the end. They have done this so well, for so long, that they have been the subject of massive parody, but also arguably inspired a lot of people to use variations on their tricks in more adult drama over the years.
Simply put the problem is that in setting up games, game companies need to actually get writers who know what they are doing, and have them involved in more than a cursory fashion in the entire process of putting the game together. Most games that tout a connection to a famous author have limited involvement by the person in question.
In creating a franchise or ongoing story, the trick is of course to have someone sit down and outline the overall plotline, how, and why things are going to be happening ahead of time.... which rarely happens other than in the most vague sense with video game series. Especially ones based around any kind of mystery or suspense.
The problem with Yahtzee's criticisms here, and in other places, is that he seems to be picking on the institution of sequels, rather than simply saying that game developers are morons who have not evolved along with the medium. 99% of the problems he points out, both here and in other areas, largely comes down to the gaming industry being resistant to change. As a number of sources over the years seem to have been saying, the actual writer of a game winds up taking a back seat to the guys actually doing the development, and/or the artwork. Either that, or the writing is being done around a certain set of gameplay mechanics, such as "we want to make a shooter" or "we want to make a puzzle game with portals" and then making a plot that works around the mechanics they want, which tends to be counter productive when it comes to writing and especially when it comes to trying to extend that story into sequels.
To be brutally honest, when it comes to say "Portal 2" in paticular, the big problem is that they took something that was awesome as a minor little extra in a package and tried to extend it into a full fledged AAA game. Portal kind of worked because on a lot of levels is was a giant tutorial, and a sort of tech demo, showing off what could be done with this kind of physics system. I think it impressed people because it was new, and showed off a possible future for gaming. What Valve SHOULD have done was take the Portal technology and used it as part of other game franchises, as an additional mechanic, as opposed to the theme of a game because of how people liked it to begin with. It's sort of like how they kind of sold people on the idea of a "physics gun" and you see it widely imitated as an idea in things like "Dead Space". The whole "Portal Gun" concept should be intergrated into other titles, rather than trying to hold down a franchise itself. The potential for using portals to line up fancy shots in shooters, or open up doorways to manipulate things with another kind of physics device is rather huge, and has a serious chance of outright creating smarter games for people.