Abstractly yes, the scientific process should eventually produce theories which never again need to be updated, and perfectly predict all future data. Science doesn't have a way to know when it has finally stumbled upon such a permanent theory though. Science isn't in the business of certifying its theories as true, just discarding them as false.

There are truths involved in science, but they are just mathematical/logical truths. Data is taken to be axiomatic, and logically true statements can be derived from the data. e.g. "95% of the values fall in this range", "no value is greater than 50". Those are just logical statements about sets, and they couldn't be used to prove any theory true, although they could prove one false.

> Science doesn't have a way to know when it has finally stumbled upon such a permanent theory though.

It doesn't need to and I didn't claim it was trying to (especially as it's impossible as you correctly point out). But the closer the model is to reality, the more useful the model is, and therefore the closer it is to being true.

> Science isn't in the business of certifying its theories as true, just discarding them as false.

Given the immensely complex and sophisticated and successful human achievements that depend on these models, such as spaceflight, computers, GPS, any number of things. I think you're doing the baby/bathwater thing. The models work very well so they must be very close approximations to what the universe does. So they approach 'the truth'.

I think we may have to agree to disagree. It may be you see as a binary thing – it is or it isn't. I see it as graduated.