science

This.

The models we have of reality are based on observations and forming theories that attempt to describe the observations.

Our models are, by definition, models and not the reality itself.

Since the paper is only based on the models and not on reality itself (which it can't be since we don't have access to the real inner workings of reality, so to say the "source code of reality"), the paper cannot actually say anything about reality, only about our understanding of it.

And pretty much any physicist worth their salt will freely admit that our models and our understanding of reality are flawed and imperfect. They are good and good enough to be used for a ton of real-world applications, but they are far from perfect and physics is far from solved.

Wow, this just made me realize. If we really live in a simulation the simulation or some parameter of it could be changed anytime.