Does this proverb sound familiar?
"Fool me once, shame on you; fool me twice, shame on me."
Maybe MongoDB is great now. That doesn't change the fact that many people have been burned by MongoDB.
But, perhaps we can let maybe we can let bygones be bygones if we can determine that MongoDB is solid and dependable now. How do we make that determination?
If we decide that MongoDB is now robust based on present popularity and general satisfaction, that's fallacious reasoning:
https://yourlogicalfallacyis.com/bandwagon
After all, there was tons of stoke about MongoDB early on, only for people to realize later that their data was being silently nuked, which is why there is such a grudge against MongoDB in the first place. If an appeal to popular belief is proof, then we have a paradox.
Not everyone has the time and skill to thoroughly analyze MongoDB like Kyle Kingsbury / Jepsen:
https://jepsen.io/analyses/mongodb-4.2.6
And even then, there are so many other databases that have a better track record, so it would be hard to argue that the effort would be worthwhile.
I don't see how you're struggling with the notion of reputation here.
Maybe bias in favor of MongoDB is too strong? If that's the case, one last ditch effort:
If your friend's dog has gruesomely bit you 10 out of the last 10 times you've visited him, but he swears the dog is now rehabilitated, do you just shrug off being bit over and over again, and visit your friend without even the slightest apprehension? Or do you, at the very least, remain hypervigilant, given you're history with the dog?
Now replace "dog" with "MongoDB", and replace "bite" with "invaluable production data being absolutely eviscerated, resulting in your customer's lost trust and also immeasurable financial harm being done to you and your family".