This now marks the 2nd time we are "letting" this company fix the same mistake, and it gets harder to assume good faith when the same issue crops up repeatedly.

Calling out poor engineering is not prohibiting them from fixing it. It's calling the bluff on "oopsie woopsie a privacy issue again, silly us," for a company whose primary KPI is public profile engagement. The more things like this happen, the less they look like mistakes.

It's taken me a long time to understand this, but sometimes the single mistake is something that leads to a cavalcade of mistakes. That could be putting a burned out person in the review pipeline who greenlights everything, it could be an overzealous engineer trying to make a name for themselves with retention and engagement, etc...

As pointed out earlier, we do need to let people make mistakes, and often their paths will appear to be non-linear. I struggle with this too, but a world that doesn't understand that is much worse.

> The more things like this happen, the less they look like mistakes.

That's a fallacy and being kind to humans is hard.