The impedance curve, as it relates to driveability, also has to do with impedance phase angle vs. frequency (greater phase angle magnitude = a more "reactive" or tougher load, lower phase angle magnitude = a more "resistive" or milder load). In combination with the impedance magnitude vs. frequency and the sensitivity these describe how difficult a speaker load is for an amp to drive (i.e., how much current the speakers demand and at what frequencies). It's the impedance magnitude curve that will give some idea about how much frequency response modification will result from using an amplifier with high source impedance, such as some tube amps. But the important thing to take away from this is that these measurements all vary with frequency, and cannot be accurately described by the "nominal" rated impedances that manufacturers spec their speakers at.
Unfortunately, these "complex" speaker measurements aren't something most folks are equipped to determine for themselves at home, so other than auditioning a given speaker with a given amp and judging the sonic results, you must rely on manufacturer's or reviewer's detailed measurements to get an informative idea about how easy or hard a speaker should be to drive and how it might modify the response of a partnering amp. (In the real world, driveability is also affected by factors ranging from types and sizes of drivers and cabinets, to room size, to preferred program material and listening levels, and the speaker cables used have their influence as well.) John Atkinson's commentary on his test measurements of loudspeakers in Stereophile is very educational about this subject, but note that the simple reference "dummy" speaker load he uses to calculate response modification in his amplifier measurements (based on their output impedance) shouldn't be taken as anything like generally representative of all speakers, as his own measurements show.
