prisoner's dilemma

lately i have thinking about thoughts we have...how many of the things seem so trivial and yet doesn't work out. in game of prisoner's dilemma if both players cooperate, payoff to each is say 4, if one cooperates and other defaults, defaulter gets 7 and cooperator -2, if both default they each get nothing. here defaulting is dominant strategy but if games are repeated it will lead to losses as both the player would default.

i came across an example in a book "Games Indians Play" by Raghunathan. he talked of an experiment involving interaction of various world dwellers in which everyone's aim is to maximize his/her own payoff. various strategies were put to challenge, a tit for a tat, a tit for two tats, continuous defaulting after a tit (Grim strategy), random defaulting amidst cooperation, etc.. players with all these strategies were made to play in an simulation environment and the winner was - a Tit for a Tat also called Gentleman Strategy. All other strategies which seemed to be smarter hurt each other so much that they failed. taken on one to one basis, this strategy seem to loose as either you would be at par with opponent or a loser, but in many player environment, opponents come to know you and realize that they can cheat you once and win but would eventually loose out on much reward, so they stop defaulting... not so with other players who are random. Grim strategy loses out as by not forgiving, chances of future benefits are curtailed. thus being a "gentleman"[i should also use "lady" here], puts you in better position than any other strategy. COOPERATE till opponent defaults, DEFAULT till the opponent start cooperating again and continue, SIMPLE and a WINNER.
this idea is straight forward and proven to work, then why people don't apply it in their interactions, maybe because they feel only they can be good, or only they are smart or maybe they don't trust or they are fine with the things as it exists... or this thought which seems so trivial isn't so trivial.