On the tangent of easy or difficult loads, the perspective from speaker amps and speakers is this:
As guru implied, the impedance of orthos tends to be more even throughout the frequency range. Take BG NeoCD planar drivers. A flat 4 ohms from 20Hz to 20kHz. Whereas with traditional dynamic drivers (especially in multi-way configuration with crossovers) can tend to have really screwy impedance with some spots at nominal rated 8 ohms, a rising impedance to 16 ohms or more in the treble, and on occasion even some dips way down to 2 ohms (getting closer to a short circuit).
Amps don't like driving things which are near a short circuit. Such amps do exist, but I believe they are called arc-welders, and are probably not suited for reproducing audio frequencies. Another important reason is that transducer systems which have screwy impedance will have FR irregularities depending upon output impedance of the amp. This is rarely a problem with solid-state systems with large amount of feedback = very low, almost 0 output impedance. Tube amps however will have a more difficult time.
Now going back to the case of multi-way dynamic speaker systems. It is indeed possible to flatten the impedance curve of such systems through use of passive components, i.e. caps, coils, resistors, etc. The downside to this approach, is that these components, while they make the speaker more palatable to amps, also act as heat-sinks essentially wasting energy output from the amp. Some designers elect to implement these impedance compensation circuits, others elect to leave them out (and assume you have a super duper mega power amp which be double as a welder.)
At any rate, orthos are different in that they tend to have lowish impedance, but poor efficiency. I would not say orthos are an easy load to drive compared to Grados / ATs (extremely efficient, low impedance) or Senns (moderately efficient, high impedance).
More thoughts later.
