I'm taking a stab to open the analysis. Based on the wording of the Wikipedia entry the puzzle is to devise a test using a test load and a multi-meter with the cover plate in place. To detect whether the ground is carrying current one would connect a load to one receptacle of a duplex pair and then on the other receptacle measure the voltage difference between the neutral and the ground or the hot and the neutral or ground. If this is a single receptacle one would use a multiplug cord or adaptor. Somehow the results would reveal reversed neutral and ground or bootleg ground.
Another possibility is with the load connected use an extension cord to a known good receptacle. Measure the voltage difference between a good ground and the ground and the neutral on the receptacle under test.
The detection of a bootleg ground in a given receptacle seems straightforward. Connect a significant test load (10 to 12 A) and measure the potential difference (voltage) between the neutral and the ground. If they are tied together in the box, the voltage will be close to 0. If the ground and neutral are tied only at the panel, the voltage will be > 1 V to ~ 3 V. I have a simple Fluke tester which will not read less than ~4 V so using that meter I would measure the voltage between the hot and the ground and the hot and the neutral. If these are equal, then there is a bootleg ground.