I really should think before I post.
We've achieved 99% photorealism in Hollywood movie special effects, while there are some impressive real-time effects in modern video games that are approaching photorealism. Will there eventually be a stopping point, where a computer generated image has the same level of detail as the real life object it is attempting to mimic? For instance, if a human character is rendered with individual hairs and pores, why stop there? Why not take it a step further and model all the way down to the cellular level? If taken beyond that level, why not also include individual moelcules, atoms, aub-atomic particles, quarks, gluons and finally strings, if they even exist? It's difficult to imagine the practicality of having that level of detail, but the implication seems to be that
a) There is no reason to believe that, as long as processing power increases, we can't render the same level of detail in a simulation as we experience in real life, even at the microscopic level, and
b) With enough processing power, we may be able to render beyond what our most powerful equipment can display to us, such as say, rendering all of the strings making up an object, or to display the entire 28 billion light-year diameter of the known universe simultaneously.
Again, the practicality of such rendering is debatable, but this is more a question of whether we can, rather than why we should.