If scientific theories are useful in explaining the universe, would it matter if they are not true?

A theory that is false tends not to be useful for long.

Theories have to accommodate incoming data, and they don’t do that very well when they’re false.

A lot of philosophers, e.g. Quine, go on and on about how different data-sets can be equally well accommodated by different theories.

And that would be true if theories modeled frozen data-sets.

But they don’t, and it isn’t.