Now it does and it was my fault that reboot and reset did not work as expected.
As I wrote in an earlier post, I am always investigating the interaction between USB QC and PD chargers and my various SBCs.
When I noticed the behaviour described, I had set up an NVMe as the root filesystem and connected a third USB power monitor, which I had ordered previously. In the course of the tests, I ordered other USB cables. With these, I destroyed the second USB power monitor on the VF2 after a shutdown. In the meantime, I ordered a fourth USB power monitor. With the latest USB power monitor and the new USB cables, the VF2 behaves completely normally.
Some may wonder if all the power supplies, cables and USB power monitors weren’t quite expensive. I can only say that yes, they were, but I am prepared to pay almost any price if it means I can learn something. But since I only really enjoy learning when I can give a talk about it afterwards, I will report in detail about my new USB-C PD experiences in the next few days.