I want to believe the same thing, but you can't really judge whether there's a defect when the number of incidents is so small. That only works if you're talking about the defect of a known single point of failure.

Imagine there are N = 3 safeguards in the product design (note: I have no idea how many there are in reality), and imagine 1 of them is defective across the whole line. That means the other 2 remaining ones would still prevent almost every incident, and you'd indeed need to be unlucky for all 3 to fail. When you observe the first failure, the only thing you can conclude is that, in your particular case, all the safeguards failed. You cannot conclude (with any useful degree of confidence) that there was no common defect across the entire product line, given there are 2 other safeguards to prevent disasters. To figure that out the only practical way (without risking more incidents) is to actually examine the product and the failure mode.

(Now, admittedly I'm oversimplifying a bit here, but the point stands. e.g. if there are 2 safeguards, each with an independent 1/10k chance of failure, and you get 2 failures within the first 20k items manufactured, then you know that's incredibly unlikely without a defect. But even this reasoning only works after the 2nd incident (not 1st), and it still requires you to know the characteristics of all the safeguards... which is not information consumers have immediately available, and neither of which is the case here AFAIK.)