I do wonder if the dog would have behaved that way if the humans were gone. Experiments seem to suggest that the majority of dogs of all breeds not trained in personal protection are less likely to engage an intruder, once they have gained access, in an aggressive manner when home alone. My responses to this issue have been based on a dog protecting while I'm gone - that's what I think of when someone says security system. Not being a deterrent (which they are, of course) or giving me time to get to a weapon while I'm home, but protecting the home and the possessions inside while I'm gone. I still believe that an untrained dog protecting the possessions in a home when their humans are absent would be the exception rather than the norm.
Depends on the intruder. If somebody is totally confident and not afraid of the dog at all usually they can call the dog's bluff. Richard Heinz has a ton of examples of this on his channel (Only mentioning him because Steven C and Marke have posted vids from him in the thread ). From all breeds. He even demonstrates what happens when you show fear with the same dogs. Then they bite or go on the attack.
But keep in mind he is a professional level dog trainer and he has on a full bite suite too. As do most decoys who test that. Or sleeves.
He says 90% of all guard dogs he tests like that can be punked out like that.
So he agrees with you. I do too on this. I think 90% Is way high. But who knows.
I just feel that most intruders wont have the same level of confidence and fearlessness as a pro in a bite suite. Or the experience to notice a bluff on a snarling, barking, growling, and lunging 145lb Cane Corso.
But this is exactly why I test my dogs. Too many variables to leave things to chance.