I recently called John Atkinson a numbskull (Mikey was not happy) because he implies that a speaker's frequency response will follow its impedance curve when used with an amplifier with low damping factor. This is not true. Speakers, even electrostatics, operate on power. Sure, put a voltmeter across the terminals of MLs and it would show decreased high frequency response. Put an (ideal) ammeter in series with the speaker it would show a rising HF response. Correct? I have done this with mine (Montis, tube amps w/ 0.7Ω output impedance, compared with solid state): put a microphone in front and measure their response and there is little or no difference.
Power in this case is factor of voltage in speaker terminals. Typical speakers and MLs too are voltage driven thus output SPL will depend on drive voltage which will be affected by output impedance of the source (damping factor & power amplifier). Voltage on speaker terminals is result of voltage division between source and load impedance.