We are reviewing Scenario 1 currently as coming out of sleep is not like a system reboot. Scenario 2 would result in Safereboot (the component responsible for the reboot dialog and managing the reboot) would resume as it left off and execute the reboot.
Hope to have an answer on Scenario 1 here for you later this week.
Scenario 1 - If a computer is asleep at the time of deployment it will miss the patch cycle. When it is woke back up it will not patch on boot or anything like that. It will continue to miss updates until it is left powered on and not put to sleep or shutdown during the time of the patch deployment.
I am hoping this is not the case. This means machines in sleep mode are pretty much unprotected.
Maybe Chris will have better news for you but from personal experience it just misses the cycle. My guess is that patch on boot is looking for a particular windows event and that is what triggers the patch on boot to happen x number of minutes after that event.
No good news on Scenario 1. Coming out of sleep will not cause the task to execute. The system does not have an event around resuming after sleep or hibernate like it does around rebooting. We would have to figure out how to capture those events and trigger the missed job.
We will look at something in our next release for this. It will take some investigation before I could say if\how we will do it.