Nick's Picks, maybe you're thinking of the way microphones with output transformers behave, as if all balanced microphones behaved that same way; your conception of the way things work applies 100% to microphones with output transformers.
But transformerless balanced microphones don't necessarily have symmetrical output circuits. They're still considered balanced if, for example, one pin isn't driven at all, but merely has passive components (such as an RC network) connecting it to circuit ground through the same impedance as the pin that's actively driven. Some very widely used transformerless professional condenser microphones--I'm thinking for example of the Neumann KM 100 and 180 series--are built that way.
Which pin they drive varies, however--the KM 100F amplifier drives one pin while the standard model KM 100 drives the other one (in opposite polarity). As a result, no matter whether you choose to ground pin 2 or pin 3 at the input to your unbalanced recording device (via a coupling capacitor of course, so that you don't short the phantom powering), with some kinds of very good microphones you will be shorting their entire audio output to ground. Whichever pin you choose, you're definitely screwed some of the time.
And that's not the only type of output configuration to contend with. The Schoeps circuit drives both pins 2 and 3 actively; if you short either one to ground (again, even if only at AC), you're loading half the circuit with a far lower impedance than it's designed to drive--and that is virtually guaranteed to cause premature overload.
I frankly don't know how DPA, AKG, or various Chinese manufacturers arrange their transformerless output circuits but I do know that each manufacturer of professional condenser microphones gives their own instructions for the proper connection of their own microphones to unbalanced inputs, and that the different manufacturers recommend at least four different schemes (that I know of) for doing this--and in some cases with one manufacturer alone, the instructions vary from model to model depending on the exact configuration of the output circuit. Choose the wrong approach for a given microphone type, and you may well get the silent treatment or early brickwalling.
So I repeat: The ONLY passive approach that works in all cases is an input transformer--and that, of course, effectively makes the input balanced.
Or you could build an active buffer circuit with a balanced input and an unbalanced output. But that can become a difficult project by the time you accomodate all different microphone sensitivities and impedances, and the whole point was to avoid active circuitry that would need its own powering, no? Otherwise, people might as well just bring along a mike-to-line preamp with balanced inputs and unbalanced outputs, and use the line inputs of the recording device.
--best regards