This is the age old question of engineering. Scientific "truths" are actually descriptive models that can (for a given circumstance) be relied on to correctly predict the observable results of an experiment. Depending on the nature of the experiment, and the precision required, multiple descriptions/interpretations may provide equally "correct" predictions of the outcome.

For most applications, a rough approximation (i.e. Newtonian physics or basic electrical circuit theory) is good enough. Some applications require more specialized models. Quantum theory seems to be very useful at predicting observable results in high energy physics (and several other areas of scientific interest) but it is a lot less useful than much simpler models in other fields.

- Eric