The whole reason speakers started to have 2 sets of terminals was not for bi-wiring, but for bi-amping. Current starvation in the output devices, increased IM in the output stage from non-linear effects at higher currents, and IM from modulation of the power supply from bass frequencies is eliminated from the amplifier that only supplies the typically much lower current mids/highs. Voltage induced distortion effects at other stages in the amplifier will still occur of course, but traditionally this has been less of an issue.
There is no technical justification for "systems with properly designed passive crossovers make poor candidates for bi-amping", since the benefits are not for the speaker, they are for improvements in amplification. For amplifiers with lower THD at high power (typically because of high distortion near the 0-crossing), the high frequency amplifier will increase in IM, but will reduce in THD by riding on the high voltage bass signal. That same amp with an input passive filter for the highs can have sonically unacceptable distortion characteristics on real music.
Systems designed to be bi-amped typically do not have internal
crossovers. Systems with properly designed passive crossovers make poor
candidates for bi-amping.
The parallel loading of the additional drivers on the amplifier is almost always negative (see above) if your goal is accuracy and un-colored sound. Bi-Wiring only adds a very small amount of isolation, basically the impedance of one set of wires, which with competent wires is very little. Bi-Amping completely electrically (and back EMF) decouples the two driver sets.
driver response varies in combination with the driving amplifier and the
parallel loads of the additional drivers.
