Isn't truth the only thing science cares about?
Of course truth alone doesn't mean much and we usually conduct science not just for the sake of it, but those are downstream concerns.
> Are Newton's laws "true"? Well, they work reliably given certain constraints and they are definitely more simple than other theories, so it would be stupid not to use them.
Science doesn't necessitate any "reliability" on the theories/premises it uses. You set up your premise apply your theory to it and reach a conclusion.
It's up to others (i.e. engineers) to worry about how valid the premises are.
Deciding to use or not use some other theory as a premise isn't about being or not being "stupid" but rather about being able to prove something that others (be it scientists or engineers) will be interested in reading.