Of course it can. Utility power drops, the "killing" relay drops, the genset starts. Some amount of time later it comes up to speed, and depending on his hairbrained wiring, the relay energizes from the generator produced power. Relays do not transfer instantaneously. It takes time for the relay contacts to move. During the time that the N/C contacts remain closed, generator power is fed out through the mains breaker to the grid. Again, that contradicts you contention that generator power is "not ever" connected to the grid. Some number of miliseconds after that the N/O contact makes, which kills the genset.
Or, if there is a defect, the relay does not kill the genset.
The point that was made addressed grid power at the service panel transitioning from off to on. You snipped the following:
***begin quote*** When the mains power is restored, it will take some period of time for the relay to move the contacts off of the N/C position. During that brief period the genset will be connected to the grid. That period, no matter how brief, contradicts "not ever". ***end quote***Transfer equipment must be 100% fail safe. It must ensure that the genset is never ever connected to the grid. Not for 1 second. Not for 1 milisecond. Not for 1 microsecond. Not for 1 nanosecond. NEVER.
His proposed circuit is not only NOT fail safe, it is highly failure prone, AND it can connect the genset, however briefly, to the grid even if the circuit has no failure, as discussed above. It fails, even with no bad components, because the design is wrong.
Ed