True, true, impedance curves will change what power is dissipated across the speaker. Standard resistors are used to level the playing field so that comparisons can be made.
If you were bored out of your mind and had a lot of time on your hands, you could plot max power vs. frequency on your speakers and see that the max power would be an inverse of your impedance curve.
Actually, your impedance curve would change suddenly as you blew your midrange and/or tweeter. The woofer might not survive such a test either.
A better way to do this would be to mimick the impedance curve by adjusting a power resistor for the frequency points on a impedance plot. Wow, that sounds like such a waste of time.
Keep in mind that the voltage output on a feedback design amplifier is the same (constant) for 4,6,or 8 ohm speakers. It's just the maximum output voltage will be different. Lower ohm loads cause more drop on wiring, suck more current out of the supply, cause a larger voltage drop across power transistors, etc. VI limiting circuits will kick in too, hoping to guard the output stage agianst dangerous conditions. This reduces the maximum voltage level at clipping. If your not going to run at clipping all day long, I don't see the need to compare output power ratings vs. load impedance. It's more important that the amp have an adequate thermal system and enough output transistors to share the output currents.