Gundato: I am not sure if I would go anywhere near that far. But if your argument is that capitalists are too damned greedy, you might want to go take a look at how economics "work" :p
Aliasalpha: Nah this is nothing to do with greed, it seems a matter of size & business attitude.
Valve were & are a comparatively small company with a very small release schedule. Unless I'm mistaken, they're only ever made 2 games, halflife 1 & 2, everything else they've released was community made mods that they bought the rights to and polished up (was left 4 dead a mod as well? I'm a bit hazy on that one). The limited releases they make mean that they have the time and motivation for fostering a community in the hopes of long term financial benefit. Counterstrike proved that modders could not only extend the life of a product causing more sales of the core game (generating mid life revenue and keeping the company going) but modders could also be bought out and become employees giving a major return on investment when they can SELL copies of counterstrike rather than just show the extra cool things you could do if you bought half life. I don't think its unreasonable to suspect that without mods for half life 1, we'd not have half life 2 (and steam) today.
Activision, by contrast, is a major corporation with a widely diversified list of games and development studios. If Game A doesn't sell as well as they hoped, its a lot less of an issue because they always have a new game in the pipeline. That sort of fire and forget attitude is less likely to foster long term community support but that isn't financially neccessary to keep the company going because they always have a fallback. If fostering a long term community though features like mod support, clan support and the like isn't neccessary for long term success and costs money to implement, there's little point in a large company doing it.
L4D involved them buying the studio that did CS: Condition Zero (I think) and paying them to develop a game.
They "bought" some random guy who made a cool Portal-like game for his term project at a university to do Portal. And they did something similar for that new game everyone keeps mentioning (a DotA-clone, I think) that I can't find any details on.
Half-Life and Blue Shift (I think) were made by "Valve proper".
Opposing Forces (best of the GldSrc, by the way :p) was by Gearbox, as evidenced by the endboss of OF being the endboss of Borderlands. Shepard was hardcore.
HL2 is "Valve proper", as are the episodes.
Pretty sure Team Fortress 2 is also largely "Valve proper", but I wouldn't be surprised if they bought a few members of that mod team too.
CS was a mod that got bought. Not sure if the team that works on it is more "Valve proper" than not though (grey area). Same with Day of Defeat (the forgotten little brother of CS, characterized by deployable machine guns and rifle grenades that people don't bitch about).
Think I covered all the Valve games.
As for the Activision approach: The thing to remember is, that "fire and forget" mentality means they can take a LOT more risks. And that is a VERY good thing.
The only game Valve ever really took a chance on was Portal, and even that was a no-brainer for being awesome (and was actually sold as an incentive with Ep2 and TF2 anyway). Everything else they make is a pretty standard First Person Shooter. Why? Because people like FPSes. Hell, one could actually argue that Valve are less creative than most of the major dev teams, since they never stray too far from their comfort zone. But when you release one game every few years, you need to make sure that game is going to be popular. Steam has given them a lot more freedom, but I still don't see any new ideas coming from them (even Portal 2 is pretty heavily inspired by all the Portal ripoffs). Does that make them any better or worse than other devs? Nope.
With the "fire and forget" mentality, you can take pretty big risks. I can't name any from Activision off the top of my head (I don't really pay attention to publishers, I generally focus on dev teams), but let's use EA. Spore. Like it or not, that was a huge risk. It was a game that really wasn't the standard PC gaming fair. And it sort of bombed (because PC gamers didn't want that kind of game). But it represents the exact thing everyone always begs for: Taking a risk to give us something new. Same with Black & White, to a lesser degree.
As an example of how NOT to do things: Introvision (or whatever they are called). They release a new game every few years, and it is always a niche game (and, in the case of Darwinia, kind of sucked...). Microsoft had to yell at them and teach them how to do proper game design with the XBOX version of Darwinia. And, as a result, the fans have had to go around and spam other forums begging people to buy games or give donations (I count at least two times :p). They don't have the capital from a "fire and forget" mentality or the Bethesda/Valve/pre-2005 Epic "make people worship us to the point that they will buy anything we put out there" approach.