Yes. That is literally the point. It is absolute proof that discretization errors exist in every equation involving PI or any other transcendental number, which utterly destroyed the trolls claim that discretization errors only exist in computer science.
Discretization errors are everywhere in math but they are rarely impactful outside of iterative algorithms (algorithms that repeatedly feed their own output back into their input). The reason discretization errors become relevant here is because the errors compound exponentially. Itâs such that in weather simulations, for example, a 1,000,000x increase in the accuracy only increases the stability of the simulation by roughly 3x.
So if you are trying to solve really simple problems, like those found in highschool physics class, then discretization errors arenât relevant. But if you are trying to solve any open modern problem, ranging from cosmic simulations of galaxy formation to deciding how water will flow inside a pipe to whether or not it will rain next week, discretization errors are an extreme challenge.
Basically smooth curves only exist in theory. They never exist in the real world. In the real world cylinders arenât perfectly round, materials arenât perfectly homogeneous. The reason mathematicians in ages past would use continuous functions is because they can approximate the real world using basic shapes like ellipses and circles, which provides answers good enough for their purposes. Well the real world isnât a nice clean set of basic shapes. To represent abstract objects you have to break them down into points, a process known as discretization, which then suffers from discretization errors. But, it allows you to represent the shape of abstract objects much more accurately which is better than approximating their behaviour using basic shapes & continuous functions.
So basically the entire realm of mathematics ranging from fluid dynamics to einsteins theory of gravity is heavily plagued by this problem. You simply canât approximate the movement of stars in a galaxy using basic shapes. The very idea of representing such a complex shape as a sphere is bonkers insane especially over the time-scales you are calculating for (more time = more error).