I can say from personal experience that if one microphone in a pair has a different sensitivity level than the other, odds are good that the response curves are going to look different as well. It's true this will not always be clearly audible in a recording of the output from a PA system at a distance. However, it can be noticeable at the time of making the recording. Behaviors of the signal that make the recorder levels seem uneven, causing adjustment, can quickly dance from left to right. This is caused when one mic is more sensitive to certain frequencies than the other. What looks like great stereo separation can actually be surges from the variations between the mics taking a particular note that should be represented in both channels equally to be louder in one mic, which distorts the imaging. This is why matched pairs are not necessarily sequential in serial numbers. When "matching" mics into stereo pairs, mfgs are seeking two units that have the same open circuit voltage, and response curves that match as closely as possible. Users can by chance obtain two single mics of the same make/model that match closely and as long as the sensitivity matches, that's usually going to be fine for PA taping, imo.