"Casper said that there have only been about 10 observations of tau neutrinos in all of human history but that he expects his team will be able to double or triple that number over the next three years."
This is not to say that it's _easy_ to detect the other kinds, you still need a large number of neutrinos and a large volume for detection. The example that always comes up is submarine communication - which has two problems - detecting a sparse and intermittent signal to get a useful bitrate out, and generating a beam of sufficient intensity to begin with, let alone a beam that is steerable!
How about astrophage? :)
Not impossible, but likely this amount of orbital lift capacity is better used for other projects.
It's one of those cases where "just" really does apply. IR remote controls work this way, using a slow bitstream to key a 40 kHz carrier that drives the IR LED. Scientific applications that need even greater sensitivity can take advantage of the fact that the expected phase of the carrier is known as well as its frequency. Devices called lock-in amplifiers are used to run a wide variety of experiments and processes using that principle.
Doing this stuff with neutrinos rather than photons, however, is one of those * * * * * exercises that the textbook authors put in as a joke.
One follow up question. When reading about low-light cameras, the number of photons per pixel seem much smaller. I guess the following factors are involved:
Several orders of magnitude reduction under low light.
Pixel area likewise much smaller than thumb.
Exposure time less than a second.
Visible light vs. all spectrum.
Well, the question: Do the numbers fit? :-)