I believe game developers are aware of this. Soldiers in most games who are hit tend to just groan and fall over, as they did in 1950s war movies, even though it is possible to depict more realistic responses (e.g., the fortunately short-lived “Soldier of Fortune” games). Also, several games seem to go out of their way to minimize the humanity of (or maximize the justification for killing) the enemy; in Doom, any human soldiers you encounter are possessed zombies; in Half-Life 1, the marines are shown killing unarmed scientists before you ever get a chance to shoot at them (although, interestingly, they arguably have their own justification, as the fate of the entire planet is at stake); in Half-Life 2, the humanoid soldiers are actually drone-like cyborgs; in F.E.A.R., most of the opposing soldiers are mindless, telepathically controlled clones. More free-form games may allow the player to, in essence, do whatever they want, and without a fully modeled society, the in-game consequences may be limited. Even in Half-Life, nothing prevents the player from killing the scientists too, if he or she wants. But my impression is that most games minimize the realism of the violence in way or another.

A further indication of the separation between in-game and real-world violence is that, in multiplayer games such as Counterstrike or Battlefield 2, a lot of people play against their friends. Perhaps they are working out deep-seated aggressions, but my sense is that players play shooter games against friends for the same reason one plays any competitive game against friends–for the shared experience, not because you’d secretly like to see them dead.

I realize my evidence is anecdotal, but as someone strongly opposed to, e.g., vigilantism, I certainly do not think there is any necessary connection between enjoying combat or shooter games and “condon[ing] or even prais[ing] real-life violence.” There may be an element of American culture that does, in fact, condone or praise a level of violence beyond that which is optimal, and that may be reflected in the level or sort of enjoyment some players get out of some games. But I’m not sure that problem can be traced even to mass media, let alone video games in particular.

The “rape role play” noted above is sort of a business model at SL, so in that sense it is consensual, but once a rape is paid for, I don’t think the purchaser has to behave as if it is a role play, if that makes any sense.

Some games feature “rape simulations” where the rape is part of the programmed play, so consenting players are not required, see e.g. http://www.honestgamers.com/systems/content.php?review_id=4775&game_id=21487&console_id=13
Whether these types of activities are available at SL I do not know.

As the Feministing post notes, there is also a game called “Rapelay” that is described here:
http://www.honestgamers.com/systems/content.php?review_id=4775&game_id=21487&console_id=13
It is a “rape simulator” so consenting players are not required.

The legal scholars who study and write about gaming don’t seem much interested in the ramifications of this, or of the rampant presence of pornography in SL.

Also, from what I read in the comments on the Gawker post, how does Second Life’s “user created” status influence how we should view this? Should the moderators of the game be able to include something in their Terms of Service that prohibits or censors the creation of such game features? And if there ever was a case similar to the ones the Professor McCann describes above (in which GTA is blamed for violent behavior), who would the defendant in that case be? The creators of Second Life, or the creators of the fantasy scenario?

I think this question extends beyond Second Life. There are many games in which you can create unique modules or use a game editor to customize the gaming experience.