Well, I had to see how it works, so I knocked up a very rough version of the Smartphone Spectrometer
http://publiclab.org/wiki/foldable-spec last night - it took about 10 minutes and cost precisely nothing - some cardboard, sticky-tape, and one sacrificial DVD-R disc. (See attached photos.)
And ...
...
It actually works!!!!
The two attached spectra were shot through my smartphone camera, but the spectra are actually much clearer when viewed by eye than these shots suggest. I was having some problems getting alignment and focus through my smartphone camera (the spectrometer was hand-held over the camera), and it wasn't helped by the very shoddy workmanship of the rough "instrument"!
Spectra from incandescent light sources such as a halogen globe are nice uniform rainbows, while a "white" LED shows three clear "sub-rainbows" for each of the RGB components, and an emission source such as a Compact Fluorescent light-bulb shows very clear emission lines.
I probably shouldn't really be surprised that it works, but it was very gratifying nonetheless.
My next step will be to make a webcam-based Desktop Spectrometer
http://publiclab.org/wiki/dsk - for which I will take my time to get better alignment etc, so that it can be calibrated in the on-line app
http://spectralworkbench.org/ . (This one might actually cost me $2 to get a suitable housing to mount it in - I already have a webcam that I will pull apart for the image sensor.) I'm not sure whether it will be able to do any astronomical service, but it's still a great little desktop experiment.