Theories that agree with existing observations but make no new testable predictions are as Richard Feynman said of string theories, “Not even wrong”.  You have to make a testable prediction to test a theory.  Until that happens what you have is not right or wrong, it is only completely useless.