Albert,
No, because I don't think that is the problem. It is simply that eventhandlers don't fire if the package is started from a checkpoint, regardless of whether that eventhandler has fired on the previous execution or not.
It would't make sense to record eventhandler completion in a checkpoint file anyway because some of them (particularly those scoped to the package) will fire many times furing package execution.
hence...big fat hairy bug in my opinion!!
-Jamie