Debugging time logic is always fun…
After the DST change, I expected my Victron YieldToday to reset at midnight.
Reality:
one MPPT at 00:05
another at 00:30
mine at 00:46 (clearly not in a hurry 
)
So no bug, just… distributed time philosophy.
Fortunately, at night solar production is zero, so it doesn’t change the nature of angels.
Still… I’d love to test this on a planet with 15-minute days 
(And just to be clear: I’m a big Victron fan — this is all in good fun!)
At least for the streetlight function, midnight is considered the middle of sunset and sunrise. I guess it will be the same for the statistics, no need to use two different midnight systems. For that to work the MPPT has to be in operation for a few days. Midnight for the MPPT will be close to solar midnight. But since it has no date or position, it cant know about DST or any other local time.
Id have to read up on it again, but i think sunset/sunrise is when PV voltage passes 10V, either falling or rising. So panel orientation or ambient light does play into it.