Actually i should be happy.
But i find it a bit annoying that after the last low level format and adapting the code to catch the crash....nothing happened so far.
Yeah, that's frustrating even if it hasn't crashed. The place where the assert happens seems very far removed from the anything SD or file related, so IMO it's very unlikely this is the direct cause (e.g. a corrupted SD card would be very unlikely to produce the observed symptoms).
Since the crash is related to a timeout, it is conceivable the SD timing could be an indirect factor, and formatting could affect that if, say the SD firmware spends a lot of time doing garbage collection or something on a heavily used card. OTOH, in normal operation I wouldn't expect OS waiting for the SD card to block whatever that event flag is related to, though CHDK could conceivable break that assumption.
Since it seems to be tied to exposure time, there could also be a seasonal element, for example, if the crash had some low probability of happening only when the exposure is >=4 seconds, it would get less frequent as the days get longer.
