Another argument says that evolution cannot possibly be tested, and what possible utilisation can there be? Well, OK, that's two for the price of one.
One way to test any scientific theory is to make a prediction based on that theory, and then see how it stacks up. Look at the nectar bat Anoura fistulata, for example. Nathan Muchhala (2007) found that this bat has a tongue that extends 89mm – 150% of its body length. He predicted that the bat would feed on flowers with a corolla (petal tube) of similar length, and then proceeded to gather data on the species of plants the bat fed on. It turned out to be the sole pollinator of Centropogon nigricans, which has a corolla 80-90mm in length. (I talked about this research in more length in another post.)
As for utilisation of evolution – the domesticated plants and animals that we rely upon are the result of selection. Darwin recognised this 'artificial' selection as a model for the action of natural selection (it's still evolution, whatever the agent), and the use of artificial selection in plant and animal breeding has been of enormous use to humanity. Or think about modern medicine's approach to drug regimes for AIDS & some of the nastier pathogens. It's well known that pathogens evolve rapidly in response to drug treatments (take MRSA, for example) – and in fact evolutionary theory predicts that this will happen, and informs the ways in which we use antibiotics to treat infections. Knowledge of evolution has led to the use of multiple-drug regimes, with each drug targeting a different facet of the pathogen, used to slow the pathogen's adaptation to to the drugs involved. This same approach is applied to cancer treatments, for cancer cells are also able to evolve resistance to drugs. A theory without utility? I don't think so!
Reference:
N. Muchhala (2007) Nectar bat stows huge tongue in its ribcage. Nature 444: 701