It helps. Thanks.
Looks like it take the Iro three days to recover from a forecasted rain event that doesn't actually occur.
At 11:15 PM the next days schedule is prepared. Irrigation is called for, but, rain forecast is determined to cover irrigation need.
Day 1: no irrigation occurs because the night before's forecast said it was going to rain this day. In fact no rain occurs.
At 11:15 PM schedule creation for next day, system bug prevents Iro from scheduling the make up irriigation to occur the next day.
Day 2: no irrigation occurs.
System bug prevents seeing that Iro will schedule irrigation next day.
At 11:15 PM schedule creation for next day, irrigation is indeed called for next day..
Day 3: Irrigation finally occurs.
I'd like to ask that you work on cutting a day out of this cycle, I.e., it takes 2, not 3 days to recover.
You'll probably have to go to twice per day schedule updates, which my guess is is not a trivial enhancement.
Flex scheduling implies real time watering. But the once every 24 hour batch generation of schedules for the next day, coupled with attempted rain forecasting, makes for the problems at the moment. I think flex scheduling has some needed design reflection.