Excellent article, I especially like the point you made about starting by building out the riskiest, hardest part that provides the most value.
It's better to fail fast if you realize what you thought was valuable is not something people care about.
I would fall into the same trap when building side-projects, where I'd start with busywork such as authentication, user management, settings, etc. procrastinating the hard/unclear stuff. That's a surefire way to get overwhelmed.