One thing I can contribute to this discussion happens to be where the subjects of planar loudspeakers and technical facts (at least those posited above by kosst) overlap. Though they are my over-all preferred design, there are very valid objections to be made against planars (as kosst has done), and reasons to not like them. That’s fine.
But there were some statements made about planars that are simply not true:
1- ESL’s and magnetic-planars should not be grouped together in terms of the load they present to the power amp. ESL’s have an impedance profile that varies wildly as a function of frequency (fancy term ;-), magnetic-planars (Magneplanars, Eminent Technology LFT’s) do not. ESL’s are an extremely capacitive load, magnetic-planars an almost purely resistive one. Consequently, ESL’s and magnetic-planars present very different challenges to power amps. That’s why Roger Sanders makes two versions of his Magtech amp---one for ESL’s, one for magnetic-planars.
2- Planars interact with the room in very different ways than do non-planars (or, more accurately, non-dipoles), but some of those ways are actually advantageous. For instance, as a result of their line source behavior, dipoles interact less with the room in terms of ceiling and floor reflections, a potentially good thing. Additionally, because of the cancellation to either side of a dipole (where the front and rear wave meet out-of-phase), planars create less side wall reflections, and the eigenmodes created by the room width dimension are less excited by a planar than by a non-planar, both again a potentially good thing.
However, the rear wave of planars presents a number of challenges to users. To prevent comb-filtering (too complicated to go into in depth here), planars need to be well away from the wall behind them. Three feet has long been considered the minimum, but that has been found to be insufficient, five feet being much better. Five feet creates a 10 millisecond delay between the front and rear wave (sound travels at roughly 1’/ms)---5’ from the rear of the speaker to the wall, 5’ from the wall back to the planar. 10ms is considered the minimum time required between two acoustic events for them to be perceived as separate events, rather than a smeared single one. The rear wave reflection itself can be dealt with either by absorption or diffusion, or a combination of both. A "too lively" room may benefit from absorption, a "too dead" one from diffusion.