Yes, our models are never complete. As someone here already wrote: "all models are wrong, some are useful".

What makes me optimistic here is that since the enlightenment we have a good track record with arriving at better and better models. In the language of David Deutsch: We are arriving at better and better explanations, where the quality of an explanation is determined by rigidity ("how hard it is to vary") and reach ("in how many situations does it apply"). His books "The Fabric of Reality" is a great book by the way, he describes how the theories of evolution, quantum physics, epistemology and the theory of computation are connected with each other. The bit about good explanations is from "The Beginning of Infinity", also a very good read.