A digital music file represents an "exact" copy of the music as it was originally recorded. If you could pipe that straight into your brain, you'd be fine. But you can't do that. You have to use speakers. Speakers are imperfect, some moreso than others. One of the ways that they're imperfect is that they can emphasize or de-emphasize certain frequencies. If you played a track that was perfect simple sound waves that increased in frequency to make a sweep of the audible range and recorded that actual volume that was produced at every instant along the way, you'd see that if you graphed those results, you'd get some sort of wiggly line. Ideally, you'd want that line to be straight, but the speakers don't produce every pitch as well as they produce every other pitch. The point of equalization is to tweak the output of the playback device before it gets to the speakers so that if the speakers produce one frequency less well than the rest, you have the playback device boost that frequency so that the speakers will end up playing it as loud as it's supposed to be. Or the other way for frequencies that the speakers overemphasize.

The end result is supposed to be that you get a reasonably stable output from the speakers, but that's not the end of the story. Some people like more bass, some more treble, etc. So you use it again to modify the output to your personal tastes. But it's much easier to do that if you start off with something that's fairly regular in the first place, thus the initial equalization process.

Oh, also note that it's not just the speakers that modify the output. It's the preamp in the playback device, the power amplifier, the speaker cabinets, the environment, etc. So it's not as if speaker manufacturers could just tell you how well their speakers respond and have you fix up your playback device to that spec. There are a lot more variables involved.