If I remember correctly, it matters a lot exactly what the noise parameter is. As soon as things get noisy enough, Grim (start off cooperating, then defect if the opponent has ever defected) starts to dominate all of the clever Tit for Tat variants. Obviously, if you make things noisy enough, then Always Defect becomes the best strategy, but Grim does well long before that.
We had an IPD tournament with noise at our university recently, and I entered a variant of Downing (essentially, model your opponent as some sort of Markovian process) which won quite convincingly (mostly because it could exploit Always Cooperate, which was in the initial pool of strategies, better than the TfT variants).
We know that Tit-for-Tat and variants do very well in iterated-Prisoner's-Dilemma tournaments. However, such tournaments are a bit unrealistic in that they give the agents instant and complete information about each other's actions. What if this signal is obscured? Suppose, for example, that if I press "Cooperate", there is a small chance that my action is reported to you as "Defect", presumably causing you to retaliate; and conversely, if I press "Defect" there is a chance that you see "Cooperate", thus letting me get away with cheating. Does this affect the optimal strategy? Does the probability of getting wrong information matter? What if it is asymmetric, ie P(observe C | actual D) != P(Observe D | actual C)?