So what I have is this very setup, but all stages that get toggled on, then off (via JMP) don't turn on the very next scan. They only turn back on every other scan.
Attached is a sample program that demonstrates this behavior. I use math and $scancounter to show that the stages I ran through each scan only execute once every 2 scans.
If you run the program, just turn on C11 to start the logic. This is a much simplified logic, but demonstrates the same behavior I am seeing in my program.
You can look at the V2000-V2060 range to see it is always 2 scans between one execution and the next. Based on what you posted, this must be a bug as it sounds like it should run through all stages every scan, rather than requiring them to be off for a full scan before being able to be turned on again.