I don't understand what it is that you believe makes a meter suitable for in-circuit testing. The in-circuit impedance is going to be affected by the impedance of components and paths in parallel to the device under test. The only thing that the meter can change is the characteristic of the stimulus signal (voltage, current, frequency, and waveform), but that is only likely to show a difference in the case where you have frequency-dependent parallel impedances or if you have semiconductors in parallel and you stay below their turn-on threshold.
The circuit typically dictates whether or not a component can be measured in-circuit regardless of the meter being used. I don't agree with you that just because two meters measure the same impedance in-circuit that it is a reliable measurement.