Well just testing in general. The real world is extraordinarily complex. Particularly when bringing in human factors (like new parents without significant prior mechanical device experience, or even trained pilots). It's impossible to test for every scenario. It's extraordinarily expensive to test for even a large portion. I'd imagine there was no test for how the stroller responds to an poorly fastened QR. Or possibly insufficient testing for how different pilots respond in a 737 Max when an AoA sensor goes bad. So the testing usually covers just enough to give a reasonable assurance of basic safety. I think that's fine. If we wanted complete safety we'd be bitching about excessive regulation and it taking 5 years and $50M in fees to bring a product to market.

I think the general approach of getting products out with a basic level of testing, then handling issues after the fact is fine.

Edit: Of course "basic" is context dependent. Clearly "basic" for the 737 Max is different than "basic" for a stroller.