First, I suppose I should define what a "theory of everything" actually is. I'd describe it as a mathematical model that predicts the behavior of any object under any given set of conditions. It should be valid in all situations, and should, experimentally, match previous observations of the universe. Some would argue that such a theory should also be "beautiful", but some definition of the word. Maybe that's true; maybe it's not. At any rate, a theory of everything should explain exactly what it claims to: everything.
Central to the idea of a theory of everything is the idea of unification. There are four fundamental forces in the universe: electromagnetism, the weak nuclear force, the strong nuclear force, and gravity. We believe that a valid theory of everything might explain how all four forces are really just manifestations of a single underlying force; this principle is called unification. At high energies, all four forces should behave the same, as components of this force. We would expect similar results when talking about the particles involved in the theory.
Let's talk about an example, a partial analog to a theory of everything: the electroweak interaction. The electromagnetic and weak nuclear forces were unified successfully by a number of theorists in the mid-20th century. Now, this unification did make some predictions - some of which you might have heard about:
- We need the Higgs boson to explain electroweak symmetry breaking - a way of saying why the forces have carrier particles with different masses (the photon is massless, while the W and Z bosons have mass). The Higgs boson was detected in 2012.
- The W and Z bosons, which mediate the weak force, must exist, as predicted by the theory. These were found in the 1980s.
- Neutral currents, a type of weak interaction, were predicted to exist, and found in 1973, shortly thereafter.
A theory of everything will predict the existence of new particles or new phenomena, typically at high energies, and would take more powerful detectors and colliders to detect them. Obviously, as technology gets better and better, more powerful particle accelerators and colliders will be built. I'm excited in particular about the International Linear Collider and the Future Circular Collider. The Superconducting Super Collider would have been amazing if it had been built, but . . . it was cancelled because of budget issues. The electroweak force provides an excellent example of predictions at high energies being verified - see, as I mentioned before, the discovery of the Higgs boson.
Now, it's also possible that we could find evidence for a particular theory of everything in nature - possibly in astrophysical experiments. To use your mention of supersymmetry (SUSY) as an example, certain superpartners are candidates for dark matter. The study of those in various environments could provide support for SUSY - although it's important to consider that supersymmetry does not imply that string theory is right, and string theory doesn't need supersymmetry. They're just close companions, and each works rather nicely with the other.
If a theory of everything keeps garnering evidence, eventually it might be accepted as generally correct, although as Secespitus pointed out, a theory can never be proven; it can only be supported by more and more evidence. Maybe we find that a theory makes correct predictions for particles with up to 10 TeV of energy, but at 20 TeV, it fails. If we find that that happens, the theory would have to modified - or thrown away.