Quit it with the fear mongering while ahead.
There are only 11 or so distinct "colors" in 802.11g and some of them are close to each other and quite hard to tell apart.
Firstly, if wi-fi would be visible, you still wouldn't be able to see it all like we can't see visible laser ray unless there is dust/water/something in the air. It should be reflected from something to be seen.
But let's suppose it's just visible per se.. Ok, colors represent waves, but they don't look like 3-5 inches, more like meters on some photos.
Secondly, signal strength should dramatically decrease over distance. In two meters it should be four times as dark as in one meter, in 10 meters it should be 100 times darker.
Thirdly, on many pictures waves don't seem to disperse properly. Looks like everybody's using a very advance ridiculously narrow-range antennas.
One of the very few things they got right from a physics perspective is your eyes respond vaguely logarithmically. Its supposedly not really, deep in the decimal points theres a buried power law factor and a buried linear factor at low levels, but close enough to log for all astronomical purposes. They worked with an astro-biologist per the article; I assume this is the "astro-" portion contribution.
Google for astronomical magnitude and there's also some interesting quantitative chemical analysis stuff having to do with light adsorption. In my misbegotten youth I spent a lot of time in a chem lab squirting weird stuff into sample tubes and hovering over an ancient spectrophotometer. A spectrophotometer is kind of like what an EE would call a RF network analyzer, but for light waves, sort of. At least that's the best EE analogy I can come up with. Eventually I spent most of my lab time daydreaming about getting home to play with electronics and computers; why was I trying to become a chemist? It was pretty interesting stuff to experience in retrospect.
The visibility is... questionable. Google for "clear air return" and WSR-88d and stuff like that. A "decent" wx radar is sensitive enough to get at least some return from turbulence. How do you know its not just ground clutter? Because the doppler shift matches the direction and speed (more or less) of ground instruments. The questionable bit is clear air return is a bazzilion dB reflection loss below actual real reflectors so it would be incredibly faint compared to actual reflectors.
Also the humorous artistic waves don't show multipath like real waves. Not unusual in an urban environment to end up multipath limited not raw sig strength limited, at least in other applications; donno if wifi is usually limited like that.
But our eyes compensate for inverse square laws, and have a logarithmic response to light.
Just try light up a big room which has no windows with a candle - you can see pretty well at the distance of your hand, but in a few meters from you - it's plain darkness. I expect wi-fi waves to be like this.
Using the grath "Perceived Brightness" at [1] we have roughly the following table:
Distance/Actual/Perceived Brightness
1/100/100
2/25/~55
4/6.25/~30
We already see light in the specific frequency peak of the sun.
If we could see wifi, it would be like a lightbulb that dimly appears to pass through walls.
Realistically the result wouldn't resemble these illustrations at all; a simple signal strength -> vibration intensity converter would be more practical.
What?
Doesn't seem all that useful from what i've read though, unless you're an electrical engineer.
If an artist pictured grass as if it reflected radio waves and not visible light waves, we'd have a interesting/weird picture but one without much value.
It wouldn't even be necessarily interesting or weird. It would only need to look like grass that has another set of lights shone on it, which is pretty pedestrian. The 'weirdness' of these images comes from the fairly arbitrary assignment of colors to these wavelengths-we-can't-normally-see. Which is utterly unrelated to the original spectra.
One could 'false color' a scene by arbitrarily shifting color assignments for normally visible wavelengths and achieve something just as visually 'interesting/weird'.
This recent EM visualization was pretty cool and a bit more substantial.
"Through a series of experiments in photographic and lighting techniques followed by customising an Android phone to act as an EMF indicator and then coding our own app in Processing we were able to visualize how these fields change over objects."
and bring a thick wallet (like, don't bother unless you've got 4 to 5 figures). And a lot of patience because the real tools have a dwarf fortress shaped learning curve.
A fine idea would be an open source-ish replacement for these type of tools. Good luck avoiding the patent minefield.
Its interesting that there are lots of really good and popular and "famous" computer algebra manipulation programs competing with mathematica type proprietary software, but not much free good, popular and "famous" finite element analysis and electromagnetic modeling type stuff.
Can't you see how juvenile this work is?