Your point about the need for prediction is relevant, but string theory is currently in the class of ideas which have no prospect of making meaningful predictio

Your point about the need for prediction is relevant, but string theory is currently in the class of ideas which have no prospect of making meaningful predictions anytime soon. You might want to argue that the mathematical formulation makes the possibility that this will change in the future more likely, but I honestly don't see how that follows.