On Theories and Abstractions - bottom up

I have been reading G.K. Chesterton lately; he writes about how the world is entirely magical, while our 'laws of nature' are not: even the simplest abstractions are derived from examples that exist around us. When we say that 2 plus 3 are always 5, and that if x is the son of y then y is the father of x, then those are conditions or relationships we observe in the real world. Once we abstract, we loose something, and our derived laws are never true representations of reality, no matter how sophisticated we make them.
I believe that this is true for all of mathematics, physics, the much debated theory of evolution, biology in general, you name it. Much can be learned from forming such abstractions, but much gets lost if we take them to be absolute truth. If the Greeks had experimented with geometry on all sorts of surfaces first, before jumping to conclusions, we may have had Riemann's and Lobachevski's discoveries millennia earlier, to give an example.
Physicists are looking for the G.U.T., the grand unified theory of everything, a formula in essence that can explain the fundamental physical forces and entities, and from there, everything else. They do not seem clear as to how close they are, but I sense optimism in their books and articles. I wonder if they have thought this all the way through. The GUT is essentially the end of physics, and sooner or later of all other natural sciences. It may take centuries to apply such a lofty formula to all processes and phenomena, but sooner or later, within a foreseeable time, the work ought to be finished. Then what? Will the GUT explain everything worthy of explanation, define once and for all which technological endeavors are reasonable and which are not, enabling great feats of engineering? Past breakthroughs have done so, giving rise to the unprecedented technological expansion of the past two centuries. But what will happen once nature's secrets have been exposed to the light and are known to everybody? There can of necessity not be another breakthrough, unless the GUT is not the GUT but something smaller, at which point we might go on and search for the super-GUT, and so on ad infinitum.
The point here is, the universe might not at all be comprehensible, or describable by scientific methods. Like a flashlight in a large cave, science may be able to elucidate only a finite area of reality at any one time, and never in all its mystical detail.