Harry Whitfield wrote:
"Sensitivity is given as 112 dB at 1 mW. I assume this means 112 dB
above the Threshold of Hearing (1pW per square metre) at 1 mW in."

That looks OK to me. My handy "Science Answer Book says:
Hearing starts at zero decibels. 10 units is a tenfold increase. The
sound made by leaves rustling is often 10 decibels. Office noise level
is typically 50 decibels. A pneumatic drill = 80 dB. A riveting machine
= 110 dB. A jet takeoff at 61 m (200 ft.) measures 120 dB. Noise above
70 decibels harms hearing. At 140 dB, noise is physically painful.

Acute hearing is sensitive. More so in some other species.

In a previous posting I may have appeared naive saying I believed GE
produced a circuit for a 2 milliwatt 2000-ohm headphone amplifier. But I
have additional evidence of the adequacy of 2 milliwatts. For many years
I worked at a radio station where we kept a pair of crystal phones near
the audio patch panel. These, because of their extremely high impedance,
could be bridged across any program line with no significant effect.
Program lines are usually adjusted to a zero VU level (1 milliwatt into
600 ohms) or 0.775 volt on program peaks. Magnetic phones produced about
the same audio output with 0.775 volts as did the crystal cans. The
difference is only in the circuit loading of the 2000-ohm magnetic
phones on a 600-ohm circuit. Point is, one milliwatt is plenty loud in
headphones. I know from listening.

Best regards, Richard Harrison, KB5WZI