Are we all missing something here? The point isn't whether a theory is true or false, it's whether it's a better fit or a worse fit. The testing is what provides information about the fit. Scientific truth is IN ALL CASES a process, not a state.

So that makes it true that no theory is provable, but theories are falsifiable to a greater or lesser extent as long as we agree about what constitutes a test.

The big difference between the scientific approach and others is that the tests are openly stated, and as PF correctly says, when there's a miss, the choices are reject the hypothesis, or think about whether the test is a valid test of what we are postulating.