Basically, there are many different audio inputs to a computer, let's look at the above example, which deals with the line-in and the normal wave volumes. In my coworkers case, he listens to the police scanner in the background while listening to other audio (movie clips, mp3s, etc). A police scanner is not constant, however. It kind of comes and goes. This program would monitor audio levels for multiple audio inputs and, if the input to a certain one passed a threshold, would increase it's volume and decrease the volume on the other inputs. Then, when the level passed below the threshold, the levels would be adjusted back to normal. In our example, as soon as the police scanner finds something and the level on the line-in goes high, the wav input would be turned down and the line-in would be turned up.
In essence, it's a weighted volume control: if line-in > 100 then wav = 20 and line-in = 150. Does this explain it any better? It's only useful if you have multiple inputs that you listen to at a time, and want to make sure you don't miss anything from one of the inputs. Ultimately, it would be kind of cool to be able to set up a hierarchy (eg, after the above if statement runs, have enter another state where this line would come up: if wav > 150 then reset normal audio levels). This way, you could prioritize what you listen to based on input level.
Just a thought...