I agree with the lessons in Worse Is Better, but I don't think that the author properly understood what he was observing. The result was a confused and confusing essay.

The way that I understand it is that what is "Good" depends on how you measure it. When we measure in terms of technical quality, we get one answer. When we measure in terms of suited to be widely adopted, we get a different answer.

We tend to idealize for technical quality, but popularity is what matters more. And once something is widely enough adopted, the technical inferiority tends to be fixable.

Shipping early and working on stability later may work for something like a video game but not for a database my system depends on thanks

I think many high profile games have over the last few years, have proven this does not work. Gamers are finally fighting back with their wallets.

If you're in nation-wide healthcare, that strategy quickly becomes unacceptable.