iPod's have generic EQ curves to accentuate or diminish certain portions of the audio spectrum based on their earphones. Vocals for example would be boosted around the 800 to 4 kHz range, rock would be heavily boosted at 100 Hz and diminish downward a dB or two for every 100 Hz thereafter until you've reached mid span around 2 kHz, then would increase increase upwards by the same amount. This is done to give a particular presence or color to various types of music. The problem is that speakers are not flat linear reproduces of sound. They also have their own inherent peaks and dips. This is due to many factors such as speaker cone material, voice coil impedance (varies at different audio frequencies,) enclosure type, and acoustic environment. Normally you would run a white or pink noise generator through your audio system in the environment it's meant to work in, then adjust your EQ for a flat response with an audio spectrum analyzer. This knocks out all the anomalous dips, peaks and acoustic resonances so that all audio frequencies are reproduced equally as they arrive to your ear. After that you would then trim the EQ to the desired effect you wish to achieve, be it heavier bass, lifted vocals, or brighter treble, etc.
Unfortunately a particular IPod EQ setting might sound great with one model of speakers and sound terrible with another brand or model (remember, it was designed to work with their earbuds.) It's like tuning a bike for performance, each setup has its own nuances and requires tuning to get it to 'run' or sound the way you'd like it to perform. Unfortunately most people and stereo shops lack the equipment to set up the audio after its installed. It's similar to installing an engine, efi, and exhaust and expecting it to perform optimally. You really need the equipment to trim out all the variables, or you could tune by ear and call it done