If anyone does a pressure test , it should be on the low pressure side. You don't break a connection anywhere. You use a test port on a shut off valve at the appliance, (1/8" pipe plug on valve or appliance valve, or T in with a test port. (You then soap the fitting you used to break into the system when done) This tests from the tank valve connection to appliance. IF there is a kitchen range on the system, simply remove one of the burners, slip the manometer hose on the oriface (they are the correct size for over an oriface) and open the burner valve to show system pressure. When you bleed off the pressure, (after the green low pressure regulator) the high side supplies the pressure until the entire system is lower then what is supplied by the low pressure regulator. You can "see" the pressure drop on a gauge when it's almost down to zero better than you can when the pressure is high. I'm referring to a Manometer showing inches of water, very minute pressures, not a gauge reading PSI. 1 PSI will peg a manometer, and it takes a lot of bleeding to get it down to 1/2 PSI, (like an open pilot for a few seconds) then you will see it drop quicker as the same volume of vapor is left out.
It's like watching water drain out of a bath tub. When it's full with a lot of water, you can't see it go down. The last little bit looks like it's draining fast. The water is draining the same speed, you just can't see it due to the volume of water. Same as pressure. You can't see the gauge drop until the very last of the pressure in the line depletes. Then it drops like a rock. That's the pressure that shows a leak on a gauge. If you play with bleeding off pressure while watching the gauge, a little at a time, you will notice it drops faster as it runs out of pressure. 10" W.C is about 1/2 PSI and below 8 " it drops fast. That's where you do your pressure test. It may seem strange to test a fitting at such a low pressure when it operates at 300#, but the "volume" of vapor at high pressure is like a cushion and won't show a decrease as it leaks, just like the water in a full tub. The higher the pressure, the more has to leak out before it's noticable on a gauge.
Temperature determines drifting speed. If the empty tank is in the sun, and your full tank was in the back of a truck getting cold air over it, the pressure differential isn't as great. After the first cylinder, the pressure is the same in both tanks. A gallon of propane liquid is going to be the same pressure as 50,000 gallons at the same temperature. So you need a temperature differential to move it from one tank to another. If the drifting slows, and you can hear it only dripping into the tank, a black plastic garbage bag over the supply tank will absorb heat causing it to boil and increase pressure to move over to the other tank. Of course no heat calls always happen at night when you're on call, and it's 15 degrees or so, so sometimes you can only drift 1/2 of a 100# cylinder. Once you equalize the pressure, it's not going to move. It's enough to get them going until morning when the bulk truck gets there. Degree day information is used by fuel companies to determine how much a customer "should" have. But the addition of a dryer, log set, or outdoor BBQ not reported to the supplier may leave a customer running out before they should. An outdoor 1000 gallon ASME tank on a zero degree night running a 100,000 BTU boiler is going to need the bulk truck if it can't keep the pressure up. They try not to let that kind of customer get below 10%. Not all servicemen on call have a CDL to drive a bobtail. I left the company just as CDL licensing was starting, so when on call, I could go to the shop and get a bobtail if I needed to fill someone in the night. Laws make it more difficult now.