I have built plenty quietly and less in the open. The open ones got better faster, and not for the usual reason.
The usual pitch is distribution. Post the work, grow an audience, the audience becomes users. That happens sometimes, slower than people say. It is not why this works.
It works because publishing forces you to write the explanation, and writing it is where you find out you did not understand your own project. Every time I try to describe a build in a paragraph, I hit a sentence I cannot finish. That sentence is always the weakest part of the design. Kept private, it stays fuzzy for months.
The second thing is speed of correction. Ship a wrong assumption in public and someone tells you within a day, often bluntly and usually correctly. Keep it private and you find out after three features are sitting on top of it.
The part most people skip is publishing failures, which is where the value actually is. A launch post with a number teaches nobody, including the person who wrote it. A note on why something did not work is useful, and you will reread it yourself.
I keep a failure log for that reason. Not as discipline. Because I kept re-deciding things I had already decided against.
Do that twenty times and you are not smarter. You just hold twenty corrections other people paid for.