The delay happens when one of the scheduled processes gets stuck and everything else is delayed. These delays are unexpected of course, and while we're trying to avoid them, sometimes they still happen. We've had long periods without them and now we have them again. The problem is that we can't really find a cause for them.
The fact that the devs are from Europe is true, and we can't cover all of the 24 hours each day. We have EGMs and other guys from different parts of the world with access to our contact info (so they can call in case of delays or emergencies) but not even that guarantees total coverage. We will give access to that info to more people to help prevent these kind of issues in the future. Also, I plan to have some kind of detection code that would let us know if delays happen. Sadly, this can only send us emails, can't really call, for that we need real people.
So, we're taking this very seriously but there are limits with what can be done, sadly.