For mic measurements you want to be using an external passive attenuator, I think (e.g. double L with 1 kOhm / 10 ohms or similar; increase to 3k3/10R or 10k/10R if needed). Cleanly generating signals as low as 1 mV may be a bit of a stretch even for an AP system. Avoiding the Pin 1 Problem also tends to be rather low on the list of priorities for a pure mic input (it should be less so for a mic/line in but that is not always the case), so a ground lift (shield disconnect) facility towards the input of the attenuator may be a handy feature as well.
Good mic inputs should be posting "shorted" noise levels in the vicinity of -133 to -130.5 dBu @20 kHz, that's 75 to 72.5 dB SNR re: 1 mV. (My own estimates have been -131.5 dBu for a Mackie 402 VLZ4 and -126 dBu for a Behringer Q1002USB.) We're really entering diminishing returns territory beyond that, as a 150 ohm resistor alone generates about -131 dBu of thermal noise at room temperature. You can push shorted EIN to -135 or -138 dBu if you really insist (which I think is roughly where the best ones are... maybe -139 dBu), but it's a lot of effort for 1-2 dB (tops) less noise.
microphone distortions are far more higher than -110dB, super super high
That'll depend quite heavily on the type of microphone and sound level.
Plain vanilla electret capsules tend to be quite crummy, alright - it is not unusual to be seeing H2 50 dB down at 80-90 dB SPL, negatively impacting speaker distortion measurements. There are ways of
linearizing them, however.
Now let's say we've got a P48 fed large diaphragm condenser mic spec'd at max 132 dB SPL @ 1% THD, -34 dBV / 1 Pa of output and noise of 18 dB(A) SPL equivalent. (Nothing special really, sub-$100 China mic performance.) The electronics usually are a single-ended affair with dominant 2nd harmonic, which allows estimating (worst-case) distortion performance by extrapolation. -40 dB @ 132 dB SPL means -60 dB at 112 dB SPL, -80 dB at 92 dB SPL, -100 dB at 72 dB SPL and -110 dB at 62 dB SPL. At 62 dB SPL, output is @-66 dBV = 0.5 mV. Now of course, at this point we're talking an SNR of 44 dB(A) (so you can turn down preamp gain quite a bit). Distortion would only equal noise around 95 dB SPL, @ -77 dB.
Now these are values that are likely to still dominate any half-decent preamp, however a great many speakers would be proud to get anywhere near them.
The Microtech Gefell M102 should be similar to what Amir's Klippel measurement rig uses - 136 dB SPL @ 0.5%, -26 dBV/Pa, noise 11 dB(A) SPL. This extrapolates to something like -70 dB at 112 dB SPL, -90 dB at 92 dB SPL and so on. It's probably a combination of smaller capsule and better-performing electronics.
Dynamic microphones have been known to sustain at least 150 dB SPL, limited by excursion around resonance - numbers elsewhere in the frequency spectrum are even higher. It is unclear to me what kind of distortion figures would be associated with levels like this, but even if it's 3 or 10% we'd still be talking something roughly on par with condenser mics.
.
