I claim tentatively that this seems to be a limitation on realism. If scientific theories are formal representations of a belief of set of beliefs about the world and beliefs are thoughts, there seems to be inbuilt limitations on how accurate our all theories can represent the world. Our brains are finite. It is composed of 100 billion interconnected neurons. Even if each neuron can be a single thought, our theories might ultimately be deficient or fine grained enough to represent the world if the world is at its most fundamental level, too fine grained, nuanced, sophisticated to be represented by such a neuronal networks. There should be a large but finite number of structures built up from basic thoughts or single neuron firings (2^100 billion(?)).
So if reality is much more nuanced in its structure than the brain is capable of forming structural isomorphisms with, there seems also to be limitations to our understanding of the world. Our theories may approach a perfectly accurate description of the world but never get there even in theory. It will have inherent limitations. If the universe is infinitely fine grained (detailed) and all that detail is interconnected in a way that it matters for an accurate description of it to represent the interconnection and more fundamental levels are required for explaining less the fundamental, then the brain will never represent reality in its essence. Even if the universe is not infinitely fine-grained (considered as gunky) or there are possible representational simplifications that retain complete accuracy but that these are sufficiently so beyond the capacities of the brain's flexibility to represent it, it will still be the case that our theories will always be strictly speaking false and that there is upper limit to how further accurate our understanding of the world can get.
My idea is predicated on a kind of representational theory of mind. The analogy (and it's a rough analogy) is with cameras. The picture resolution of a digital camera is dependent on the number of pixels it fits on a picture. The picture's resolution is limited by that number. Reality is presumably continuous but if there is enough pixels, a picture can approach a representation of some parts of the world (the scene represented in the photo). But there may be things in the world the camera cannot accurately represent due to limitations in the resolution power (for example, very fine print at some distance) and lack of pixels it can fit in a picture. The brain has analogous limitations assuming the assumptions mentioned above is true.
