He emphasizes waning motivation as one of they key factors, but there's a logical consequence of this which he didn't point out.
Instead of trying to keep your motivation up, figure out how long your motivation tends to last for (mine is ~2 weeks). Then at the very start of the project ruthlessly cut down the scope to something you think you could actually get done in that time period.
Perhaps this will require you to simplify some of the stuff you were planning to develop
Maybe you'll adopt just one new tool you need to ramp up on instead of five.
But by embracing the ticking motivation deadline you can modify the work you take on to something that you'll actually finish.
I wrote about this just a few days ago, and how I applied this technique to a project I was working on.