You are right. Weighting the test case is easy; whereas establishing the weight that the immersed case adds to the vessel of water is trickier
I had two problems. First, I couldn't get my new scale to measure the vessel with the water in it without going into overload. That forced me to get my old scale working - which had been fussy after I dropped it on the ground a few weeks ago. With the old scale working again, I tried suspending the case in the water - without touching the bottom or the sides and totally immersed.
I got a bunch of readings that were wrong and eventually found the problem was that I had suspended the case by a bunch of wire bag ties that displaced too much water. Once I switched to hanging the case from a thread - as shown in the video - it worked fine.
I got a reading of 118.8 grains for a partial WW-Super case. This was a
partial because I had sectioned it earlier. This actually help the job of getting an accurate reading of the immersed weight, because there was really no chance of air bubbles remaining inside - which certainly would have thrown-out results. The immersed weight came out at 14.0 grains. The ratio is 8.485 - which is the specific gravity of that case.
When we compare that to the table it equates to about 75% Copper and 25% Zinc - which means that the brass is below the quality of cartridge brass. I'm pretty sure if I did it again, I'd get slight differences - so I'd recommend that people do a few separate immersion tests and average the results.
Someone before said that he got a calculated SG reading of 4. If you are using a thread and get a crazy SG number that is too high, look out for the possibility that the case is touching the bottom or the sides or otherwise isn't fully, freely suspended and completely covered in water. If you get an error where the SG is way too low to make sense that could well be because of trapped bubbles. In that case, you could consider drilling out the primer hole - a lot - so that you can be sure that there isn't any air inside. The test doesn't case, if you modify the case - as long as the case has the same properties for both parts of the test - because its the ratio of the two parts of the test that counts.
To the guy who said that he was certain that every cartridge case on the planet is made with cartridge quality brass, I have to say
"... test for yourself. You might be surprised".