there's a wealth of information on "impedance" if you want to read reams, after doing a search (especially in the amp section).
if you disregard the components of the amp itself (metal oxide resistors, polyproplelyne capacitors (among many), and the types of transistors, and the amplification method (Class A, AB, B, D, H, etc), then basically you are left with a voice coil that has to produce a sound wave, or note, over a certain range. you look at it's effeciency and power handling and figure that for identical specs, they should both should sound the same. but they obviously do not. why? because most tests are probably done at 1 KHz and at different distances away from the driver. this will not tell us what it will sound like at 50 Hz or 15KHz. the driver design is then the driving factor. <pun intended>. can the driver handle a lot of power (current & voltage)?
in speakers, the highest impedance is usually at the speaker resonance frequency. for whatever reason, a note in that range will need the largest amount of power to produce (or is it the least?). now try to reproduce a fast note, like a triangle, bell, or high hat. (this plays into vynil and CD compression, also), what happens? to reproduce that fast transient note, a lot of power is necessary, usually in excess of 1 watt.
now, since the amps are supposed to state their power at 120 ohms, it stands to reason that the amp can drive a low impedance headphone to more relative volume than a high impedance headphone. So you would need a higher rated amp to drive, say, 300 ohms. (I'm talking simplistic E=IR, where E, the voltage output is constant for whatever amp. You could therefore say that there is more current available when you use a lower impedance headphone.) The problem comes in when there is a very fast transient. The voltage has to spike up, past the amp's ability to render a clear sound without it clipping. A better amp will be able to handle that note much easier if it has discrete transistors which can open and close their gates as fast as possible, without any overshoot.
in the case of a low impedance headphone, as you turn up the volume knob, the perceived change in volume is much faster than in a high impedance headphone. simplistically speaking, you'll have to turn up the knob to a lesser position than if it is a high impedance headphone. this can be a problem with linear pots which may not track well at low volumes, so you could be able to detect a channel being louder than another.
now, remember that at lowest volume, the input signal (to the amp stage) is basically shorted to ground in a potentiometer. just a little turn of the knob will produce a change in volume level (especially if it is linear). in linear, at 1/2 of the knob travel, you should have 1/2 the pot's rated resistance; at 1/4 travel, it should be 1/4 the rated resistance, etc. A logrithmic pot on the other hand goes from most (rated) resistance to 0 resistance in a more controlled fashion. So, on a log pot, 3/4 volume may only be 1/2 resistance. you therefore have finer control of the volume level.
now both low impedance and high impedance headphones are made to certain effeciency and power handling levels. mate a low impedance headphone to a log pot amp, and you'll have better control of the volume level. unfortunately, most pots today are made in the linear fashion. i believe. they are much easier to make, and therefore cheaper. cheaper means more profit for the manufacturer. ICs are cheaper than transistors; transistors are cheaper than tubes. tubes need a higher voltage, which means bigger transformers and larger voltage handling caps. but with the shrinkage of the size, there is a corresponding decrease in the signal wire thickness. this leads to all sorts of problems, the least of which is power handling.
so as amplifiers have evolved, voice coil designs have changed to be able to drive them with battery power supplies. everything has gotten lighter inside the voice coil. amps therefore are no longer rated at 120 ohms. so for the most part, impedance is not really associated with sound, provided that the voice coil has been designed correctly. so no matter what the impedance is, you'll still have to listen to it and see if you like it. <look at the frquency response graph thread>.
i still feel that low impedance headphones tend to go into distortion faster than high impedance headphones, by virtue of the fact that the amplifier tends to go into saturation at a faster pace than if it was being driven by higher impedances. yes, the volume knob has to be turned much higher, to produce the same power level, and thereby the same Sound Pressure Level, but the finer adjustment allows one to more slowly go into distortion. I say that there is more power "headroom" with a high impedance headphone than with a lower impedance headphone. This "headroom" translates into more "air" and better transients. Many times the distortion that you hear is not the headphone, but rather the amplifier saturating and compressing the sound. Since most people who buy high imp0edance headphones also tend to instinctly know that they will need more power to drive them, they will drive them with better amps. Therefore they will sound better right out of the gate.
The CD3000 is in a class all by itself.
now, about the inductance of the cable, whereby the capacitance of the cable plays a role in transients....
