Jeff Liebermann wrote:
On Thu, 16 Nov 2017 04:29:12 -0800 (PST), Phil Allison
wrote:

** By all accounts, bouncing a IR laser off a widow is not
practical. The beam has to be precisely aligned *square on*
to the window and the resulting sound quality is abysmal.

This might help:
"Laser Bounce Listening Device"
http://gbppr.dyndns.org/~gbpprorg/mil/laserl/index.html
See Fig 9 and associated text, which discusses the angles of incidence
and reflection.

In my never humble opinion, there are several reasons why it is better
to use a laser source and detector at the same location.
1. Any common mode vibration of the laser and detector mounting would
cancel if they are mounted on a common surface.
2. It's much easier to build a small interferometer than one with a
large baseline.
3. Coated

On the other foot, there is a good reason to use a large reflection
angle:
1. Reflections are limited by Brewster's angle. Beyond some angle,
al the light is reflected. Below this angle, little is reflected and
the laser beam ends up going through the glass and bouncing around the
room.
2. Since the reflected light becomes polarized, a polarized filter
will reduce optical "noise" pickup from the sun and other sources of
light pollution.

I've tried this a few time, but never bothered to try it with dual or
triple pane windows. I don't think it will make much difference
because I can hear street noises through my double pane windows with
little difference over a nearby single pane window. Both panes
probably move together.

OTOH this simple Soviet invention worked well and had no such issues:
https://en.wikipedia.org/wiki/The_Thing_(listening_device)

Yeah, that was really clever. However, I'm told that RF reflections
off of anything moving and metallic in the room made listening
difficult. At 330MHz, the transmit antenna beamwidth would have
covered the entire room and possibly much of the building. Later
models worked at microwave frequencies, which offered a narrower
beamwidth.

That is a good question: can ground-penetrating-emitted waves be bent or corrupted easier than waves within a laser beam?

Anywho, with a hard-wire connection (on the other hand) like with high-speed internet, isn't interference more easily detected (and less possible to apply)? Here, electrodes would have to be applied to the window or some other part of both the transmitting and receiving locations. Electrodes (I guess) like the ones doctors use to attach to the skull to determine brain signals, like if they wanted to merely create the sensation of smoking, drinking, doing drugs, etc..) without it actually being done.