It is a few years now since I did the AM2 and this wouldn't have been possible when I did it, as fra as I can remember the fault finding rig wasn't connected to a source of supply. All of the fault finding was carried out dead and the Martindale bell set provided was all the test equipment required for it.
Also I think it would be better to check the phase sequence rather than the rotation in this scenario. We already know that there is a phase rotation problem as the motor turns the wrong way, what is not known is which two phases have been swapped.
A phase rotation test will only reveal where the swap has occurred, it won't reveal which two phases have been swapped. You would still need to follow it with a phase sequence check.
Plus of course they probably prefer it if you avoid live testing when a suitable dead test is possible.