From the web page quoted:
The design is a standard non-inverting design, the input is applied to the non-inverting input of the op-amp which is pin 3 in most cases. The input impedance is 23.5k, the overall voltage gain is determined by R2 and R1according to the following formula:
The data sheet from On-Semi calls out a minimum of +/- 3V so you would need a larger supply voltage and buck it down to the RPi. You might have to play with the gain (R1 and R2) depending on the mic you choose. Further, the frequency response is excellent for normal audio work but may fall well short of your needs for the bats. I don't see that as a realistic goal for such an inexpensive system as you are putting together.
As far as bats are concerned, the upper end of bats frequency range is well in excess of even the 192 KHz sampling rate. The most you would be able to reconstruct would be 95 KHz according to Nyquist and even that would be a stretch and assumes a sinusoidal wave form. I am not sure about the range of the mics...you might have to get pretty costly transducers to "hear" that high. I have not investigated that at all.
As far as the zero is concerned, for me it is a non-starter for its slowness. I was able to load a full-up copy of Audacity with the 3B and 2B before that but the Zero could not access it in the Raspbian repository. When I tried to play a .wav file recorded at 48000 samples, it never seemed to play but with the 3B, the same file loaded to Audacity from the file manager and played very nicely...certainly as good as the inexpensive ear buds I use could hear. Waiting around for the zero is much like watching grass grow. A 3B+ would be a nice choice with its higher speed, on-board wifi, on-board ethernet and USB ports. Still, the power consumption is high for a remote device.