How should we measure the success of science, and is scientific success a guide to truth?
Success of science is measured by the correctness of scientific explanations.
Every scientific theory has to provide a link to observation or to experiment. The success is measured by the degree to which the theory predicts or retrodicts observed phenomena.
Nevertheless, even successful prediction of many cases does not assure that the theory is correct. Because a theory comprises infinitely many cases while all observation covers only finitely many cases. As a consequence, one can never prove the correctness of a scientific theory. One can only refute wrong theories by counter-examples. But in the case that one cannot falsify the theory, the question whether the theory is true has to be left open.
Broadly speaking, as an answer to your question: Scientific success is no guarantee for truth, but the only guide we have to make progress.
See Popper's principle of fallibilism: The Problem of Demarcation in http://plato.stanford.edu/entries/popper/#ProDem
The difference between science and other fields of inquiry lays in science's predictive power. Many disciplines: Science, philosophy, history, sociology, etc...can explain things after they occur. Sociologists can explain to us why the hippies arose in the 1960s, and historians can explain to us why the Industrial Revolution happened in 19th century Europe. But science is the only field which allows to make accurate predictions about the future, not just the past.
The text book case of a scientific theory with predictive power is Einstein's theory of relativity: Einstein predicted the existence of gravitational waves in 1916, and their existence was confirmed in 2016.
This idea was formalized by philosopher of science Karl Popper in the idea of falsifiability. A theory or statement is scientific if we can perform an experiment which proves it wrong, i.e. it can be falsified. You need to be careful with the language used here: We don't want to prove it wrong, we want to show that we can prove it wrong.
The reason is that some statements can't be proven wrong. Consider the following statement: "John fell sick because God hates him", you might try to answer by saying that "No John fell sick because he caught a dangerous virus", and your interlocutor will reply "That is true, but it was God who made him catch the virus".
There is no way we can design an experiment that proves "John fell sick because God hates him" wrong, the evidence can always be interpreted to support that statement, and hence the statement isn't falsifiable.
Compare that to the statement "John fell sick because he caught the flu virus": We can perform an experiment to confirm that John did indeed catch the flu virus, alternatively we could also perform an experiment and show that he didn't catch any virus, but that he got sick because he ingested a poisonous chemical instead. Hence the statement "John fell sick because he caught the flu virus" is falsifiable.
Falsifiability also provides a measure of how successful a scientific theory is: A theory is good as long it hasn't been falsified. Once it has been falsified, then we know that new theory is needed, and that the domain of application of the old theory is limited. This is what happened with Newtonian mechanics: It is a good theory, but it is limited to a restricted domain, and these limits were proven by falsification. Relativity is a more successful theory because it covers a wider domain, and this was arrived by falsifying Newtonian mechanics.