I filmed the operation of the 2nd loop at DHS. Security wouldn't let me stand on a bench, so I had to hold my phone above my head the whole time, which made it a bouncy video (my phone doesn't have image stabilization.)
Anyway, the operation is as expected. There are 3 positions on the loop; call them A, B, & C. 2 cabins sit in the loop at A & B, presumably A is unloading & B is loading. When ready, they move ahead simultaneously to B & C, C being just short of the switch onto the main loop. A few seconds later, a 3rd cabin moves off the main loop into position A, while the cabin in C takes it's place on the main loop. So we're back to cabins at A & B, where they remain until the cycle repeats. The departing cabin appears to always pause at position C for a few seconds, waiting for the next cabin to enter the 2nd loop.
So at the start of the video, you can see 2 cabins in the loop, one at the bottom of the loop (B), and another mostly hiding behind a pole at A. Both start moving @0:04. Position B is at a bad angle and partly blocked, but the 2nd cabin appears to stop @0:20, and the lead cabin stops at C @0:27. A 3rd cabin enters the loop @0:29, and the lead cabin starts moving onto the main track @0:33. The 3rd cabin reached position A @0:40.
6 cabins pass by on the main loop, before the cabins in the 2nd loop start advancing @1:32. 2 more cabins pass by before the 2nd cabin stops at C @1:56. The 2nd cabin departs C onto the mainline @2:01, with it's doors closed.
So during this test, after a cabin entered the 2nd loop, the 9th cabin following was the next to enter the loop. The 2nd cabin was stationary at the load position for 72 seconds. The 3rd cabin was stationary for unload for 52 seconds.