Here, he brings up many issues; I was going to start shooting some down about halfway through, but the paragraph on deliberate, focused effort lead me not to — they’re all on point.
About the only thing to add is the following: not every project is a company. If you can’t find deep value in what you’re doing, and monetize it to capture enough of that value to grow the way you’d like to, it’s a good project, but a crap company. Look at a lot of open-source software: incredible, innovative and technically superior projects that need to exist, but make for poor companies (no, I don’t mean the kernel).
If you’re exploring projects to turn into companies, make sure that the projects you’re looking at can actually become companies. How can you tell the difference? See if there’s a market!