Post History

The voltages from the raw microphone are quite small and fairly high impedance.  That's why an amplifier at the mic is needed in the first place.  The signal on the wire would be way too small compared to noise in normal environments.  The FET is a simple amplifier that increases the voltage and decreases the impedance to make it feasible to send the signal on a long (compared to the mm or so between the mic and the FET) cable.

The simplest and cheapest thing for the microphone manufacturer to do is to add the FET, make sure the gate can't drift to a non-zero DC voltage, and make the rest your problem.  Actually it's not much of a problem.  They could integrate the drain resistor, but then you'd have to run three wires to the whole microphone.  By leaving the drain open, you have to put the pullup resistor in your circuit, but only need to run two wires to the microphone assembly.

The FET is designed to pass a small current when the gate is held at 0 V.  The actual gate voltage variations are so tiny that the FET is essentially always at this simple bias point.  The AC signal with a typical 3.3 kΩ drain resistance is still only a few mV.  Since the AC variations from the bias point are so tiny, almost any bias point will do.

<blockquote>2-terminal configuration is more sensitive at low sound pressure level</blockquote>

That makes no sense.  The same circuit is used both ways.  The only difference is which end of the cable the pullup resistor is on.

With the pullup integrated into the microphone assembly, which ends up requiring a 3-wire cable, the manufacturer has more control over the parameters.  It also gives the manufacturer the option to trim or customize the pullup to compensate for the inevitable variations from FET to FET.  That could allow them to specify a more predictable gain, possibly better linearity, and the like.  I don't know how much of that is actually done.