Now
this is interesting; there's some further discussion
here.
The discovery that Tit-for-Tat is a very successful strategy at Iterated Prisoner's Dilemma was a comforting one. Even though other strategies can defeat it under certain circumstances, it is somehow satisfying to know that an essentially benevolent (though not spineless) way of dealing with the world can win over more complex, more aggressive, and less forgiving strategies. That is what makes the success of the strategy described in the articles above so disturbing. In qualitative terms, it suggests that a highly stratified and xenophobic society can provide its leaders with better results than the egalitarian society promoted by Tit-for-Tat.
Of course, Axelrod-style tournaments are vast simplifications of social interaction, but there's a fair amount of evidence that they contain a kernel of truth. One of the commenters on the second article pooh-poohs the significance of the new strategies, saying that their inventors have merely worked out a way of "gaming the system". I do not find this comforting, since gaming the system is common enough real-world behavior.
It will be interesting to see what happens in the next tournament - especially whether someone can come up with a nemesis for these strategies.