Let's see, we've predicted things like drug and poison resistant bacteria, insects and other pests, and plants. Predicted things like endogenous retrogenes in related species and their relative mutational rates. Predicted changes in physiology like the growing lactose tolerance of certain populations of humans. Predicted fossil order in geological strata and where certain fossils should be found - for example, Darwin himself predicted that humans arose in Africa based upon human and ape homologies, long before any homind fossils had been discovered. Lots of things like that.

Let's also keep in mind that physical science like quantum mechanics are fundamentally different from evolution and biology. It should be, and it is, much easier to make predictions and models based upon purely physical and unchanging parameters. We don't have that luxory in biology where environments and organisms are in a constant state of flux. In many senses evolution can be more of a historical science like archaeology or astronomy and cosmology.

Future predictions are only one small measure of a science's validity and usefulness.

Exactly. In computational biology, evolution is used all the time to construct genetic algorithms and solve problems. We can use natural selection principles to predict certain things like bacterial strain resistance and epedimiology, but the real usefulness of evolution comes in my work, protein folding. See this link for details:

http://www.rationalresponders.com/forum/sapient/atheist_vs_theist/7918

And he answered it. What you quoted is a way of saying "But that's not all" 

A more commonly seen example is "postdiction" as a kind of prediction. Evolution "predicts" the way fossils should stack up in the geological column. And what we should expect, assuming evolution, is exactly what is found.

The point here is that prediction in science doesn't necessarily mean prognostication of actual future events.