Zs via measurement is a live test, so if I'm testing a plug socket Zs - so once the Live current reaches the MET all Earthed and Bonded metal work will become live - seems a bit dangerous - what is the current and voltage the meter uses for the test?
If the Earth is connected to the MET reliably then you get current flowing but the increase in local potential
should be negligible in the test.
But clearly if you had a situation where the local MET was not well earthed, say missing wire or badly corroded terminals, or even with an earth rod for TT that could well have an Ra of 200 Ohms and be perfectly acceptable, then there is a very real risk of the local earth/CPC arrangement becoming live w.r.t. the true Earth to a dangerous potential.
It is possible that modern MFT have some safety features included, but
I don't know for sure. For example, if you were diverting your Zs test current of a few A from L to E and you observe E going to many volts above N, this is not good, and you could reduce the test current and still get a measurable Zs result.
However, what I do know is the "no trip" type of Zs test only uses a small current (probably 15mA) so it won't trip an RCD and then correlates the cycling of that on/off over a long period to see the resulting mV (or less) change in supply voltage. In this case you are not at real risk (though getting a pulsing 15mA shock won't be nice) but the test is less accurate than a high current Zs that generates a more easily measured change in voltages to deduce the impedance.
[automerge]1589706640[/automerge]
More likely a MFT would do a low-current test first, and if that caused very little change in potential difference, then up the test current.