It is often said that string theory is "not even wrong," i.e., that it makes no predictions. Brian Greene is one of the most well-known champions of string theory. A few years ago, Greene argued that string theory can make predictions; it's just that we cannot test these predictions because of our technological limitations.
So it seems that we need to distinguish between a theory's ability to make predictions and our inability to test those predictions because of technological limitations. In other words, making predictions is a formal property of a theory regardless of whether or not we have the technology to test those predictions, which is our limitation, not the theory's.
Now, if Greene is right, and string theory does make predictions, but we cannot currently test those predictions, then does that mean that there is nothing intrinsically wrong with string theory as a theory? After all, the fact that we cannot test the predictions of string theory now is simply a contingent fact about us, not the theory. We may be able to test the predictions of string theory sometime in the future.
Given this distinction between in-principle testable predictions (i.e., predictions that can be tested in principle but not in practice because of technological limitations) and in-practice testable predictions (i.e., predictions that can tested in practice), is it a legitimate objection against a theory, such as string theory, to say that it doesn't make in-practice testable predictions? On the other hand, is it reasonable to suspend judgment about a theory that makes only in-principle testable predictions, but not in-practice testable predictions?
More recently, Greene has argued that string theory actually makes in-practice testable predictions. According to Greene, string theory can predict the amount of dark energy in our universe, as opposed to other universes in the "multiverse," which is something that can be measured.
If Greene is right about string theory predicting the amount of dark energy in our universe, which can be measured, does that mean that string theory is a scientific theory?