If we assume that is one of the connectors in the trunk to the 3rd high mount brake light then IF the other end of the test light is connected to a ground THEN touching the probe of the test light to a GROUND isn't supposed to cause the test light to light.
If the desire is to test a ground THEN the other end of the test light needs to be connected to a known good POWER source.
Should we assume that the test light does light up when touched to the RD/YL wire (test lead grounded)? IF so that proves power to the RD/YL wire. To test the ground put the test light lead on the RD/YL wire and touch the probe to the BLACK wire. IF the ground is good, then the test light will light.
Bottom line. What the test is can test for depends upon what the test light lead is connected to.
Test for power: Connect test light lead to a known good ground. Probe connection to be tested.
Test for ground: Connect the test light lead to a known good power. Probe connection to be tested.
Note for advanced electronics understanding. You will be able to measure voltage with a VOM meter on the black wire because the VOM is "seeing" the voltage through the lamps. Since there's no current flow (and hence no voltage drop) this is to be expected. The likely take away from all of the is proof that the ground path isn't any good.
What more proof? Take a long piece of scrap wire and connect to KNOWN good ground. Touch the other end to the BLACK wire in the photo. Does the 3rd high mount brake light work now? If so, this PROVES the problem is a missing ground path.