My colleagues and I have discussed this at length. One aspect of dropping IE support that "piques our intellectual curiosity" is thinking of it as game theory. It mightily resembles one of the four games centered on Prisoner's Dilemma (along with Moose Hunt, Chicken, and one other that eludes my recollection).
If some critical mass of vendors/web sites/providers drops IE support, we all win. But if we don't get critical mass, those that drop it lose business to those that maintain support for it.
I like your game theory interpretation, but I think that there are subtle differences that rule out using the prisoner's dilemma as a model. A really good strategy in the prisoner's dilemma is Tit-for-Tat. So the logic goes that if your competitor drops IE6 support then you should too. But here's the trick. Since you may have multiple competitors and (I think I'm right about this) only one iteration of play, there's no really good strategy to follow; so it's not really quite the prisoner's dilemma.
Most importantly, unlike the prisoners' dilemma, you can communicate with your competitors. Perhaps what you can do is collude with them to drop the support simultaneously since it would save you both money and aggravation. You might want to keep a backup site that does have support enabled in case they don't all hold up to their promise so that you don't get burned. I don't imagine that anyone would agree to a contract on something like this, so plan B is essential.
PD is not the only game in the family, there are four games that differ in the relationship between the payoffs. Which game it might resemble depends on how you evaluate the payoff matrix.
For example, in "Chicken," coöperating means turning aside and defecting means driving straight. The payoff is very big for the person who defects while the other player coöperates, but there is a big penalty if both players defect simultaneously. Same game mechanics, different payoff table.
But all of the games have the property of having two basic choices: coöperate and defect. In this case, coöperating is dropping IE support and defecting is maintaining IE support.
Of course, these are n-player games and the matrixes are more complex because the payoffs depend on how many "players" defect.
I don't think communication matters for these games. It's always explained that you can't communicate, but teh really essential deal here is that you have to make your decisions simultaneously, no player sees the other player's decisions before making their own decision.
In that respect, real life is different: there can be early stage players like 37 signals that get "opinionated" and laggards can watch what they do before deciding.
Thanks. Most diacritics in English are on loanwords, but I like using the diacresis to disambiguate long vowels. A particularly esoteric one is Oölogy, the study of eggs :-)
If some critical mass of vendors/web sites/providers drops IE support, we all win. But if we don't get critical mass, those that drop it lose business to those that maintain support for it.