Just a note:
To tap a fiber without interrupting the flow of data,
the method is to first remove the polymer sheath
(chemistry or micromachining is your friend), and then
very carefully score the side of the glass fiber itself.
You'd have to use a diamond-edged something-or-other and
be inhumanly precise about it; most likely a machine of
some kind set up to make a score mark about one twentieth
the depth of a human hair.
This will degrade the signal in the fiber, slightly;
possibly not by enough to disturb the next repeater.
Now you've got a fiber with a score mark in it, and the
irregularity will cause light to scatter around the score
mark. Now your problem is that the light is going in
many different directions away from the scatter point.
You can use a reflector to collect and concentrate them,
and some kind of lens with a continuously variable index
of refraction to realign them into a straight path. Then
you can still feed over half of the scatter radiation
into another fiber.
Now if you stick a detector on the end of that fiber
that can decode pulses with only about 1% of the
power of a regular fiber optic signal, you are in
business.
Caveat; the whole operation has to take place under
clean-room conditions. Not gonna be easy.
Second caveat: This whole thing is sheer speculation.
I work a little with glass, know some of its properties -
but this is way out there and I'm guessing. Also the lens
would have to be a special piece of work -- I picture an
all-semester project for a materials-science/optics team
there.
Bear
