Comment by IMTDb

It’s also a biased view. The author admit that the feature he was involved in took longer to ship initially. Depending on the environment this can be an anti-pattern; don’t we say “release early, release often”.

In the same vein; the author says that the other feature took several releases to be stable. Were the other release purely bug fixes or did that give the engineer a chance to get early feedback and incorporate that into the feature ?

It’s clear that the author prefers a slow and careful approach, and he judges “success” and “failure” by that metric. It sometimes is the right approach. It sometimes isn’t.