If there was one single, coherent wifi source transmitting a continous, wide-band signal covering the WiFi spectrum, one would assume that peaks and dips were caused by wave phenomenon such as reflection, absorbtion, transmission and refraction by buildings, ground and the atmosphere causing fluctuating received power levels over time/space. At 2.4GHz, one wavelength is small enough that small displacements can cause large changes in the signal.
If they are measuring the total received power from a mixture of sources distributed in space and time in a complex manner, one would assume the signal to be... complex.
Nice concept! Photography is all about capturing electro-magnetic waves, mostly indirect (e.g. light reflected by some object). Imagine imaging (seeing) the landscape at approximately 2.4GHz the way that we photograph (see) visible light. Either passively (illuminated by WiFi transmitters and all kind of other unlicenced sources), or actively (flashing the scene with a radar-like device). Might even blend the two at appropriate weighting like we de with natural light + flash :-)
-h