Amazing how that works.
Actually, since this is by code a "separately derived system" the ground and the neutral should be bonded at this point-either in the transformer or in the disconnecting means/panel after the transformer. If that is done correctly, voltages will read correctly. If not, neutral to phase and phase to phase will read correctly, but voltages from ground to neutral or either "phase" will float and you will not have a proper path to trip breakers in the event of a fault. I'm sure there are plenty of master electricians out there that have never connected a transformer like this-depending on the type of jobs and area its not a very common task-I think I've done 3 in 20 years. But they all should know enough to meter their work to verify before walking away.
"Measure with a load connected". Do the math to make sure the measured current is also the calculated current; explain the difference, if any.
Got a phone call several years ago from a buddy who was having rising voltage between ground and neutral over distance (most of you already know what's up) so I had him take several voltage and current measurements over about 20 minutes (4 phone calls). In the end we determined that the hired generator was not attached to an earth ground, nor was the generator generator neutral bonded to equipment grounding connector. In "prairie ghetto" fashion a set of jumper cables was used to test the hypothesis, and The Electrician (Or Someone Like Him)* was summoned to install proper jumpers.
* Fire Sign Theater references for the Illumi-sceti. Shoes for Industry!