The ideas from Getting Things Done lie at the center of how I get work done during those times when I become the bottleneck. Of course, I try very hard not to let that happen to me too often.
So as not to let the book itself become the bottleneck, you can start extracting value from the ideas by reading this short guide. I give this to almost every team I work with now, and a few have called it both the single most valuable thing I've taught them and the one thing that sticks.
At least one reader has told me that the advice to write only a single next action per project has really helped him. I wish I'd come up with the idea, but I took it directly from David Allen's book. He describes why many particularly effective, motivated people find that the advice in Getting Things Done weighs them down and even causes some mild depression. This happens because they build an accurate picture of how much they want to do, then realise that they couldn't possibly do it all. That was me. This literally sent me to bed for a week.
This explains why I strongly recommend not always trying to figure out and record all the actions in a project. Of course, when you feel ready to do the project in earnest, moving it to "the front burner", then you'll probably want to explore the work in more detail, but even then, you don't need to map out every detail, especially if you tend to feel overwhelmed by the amount of work that would go into that project. If you know that planning too much will stop you from completing the next action, then don't bother. Yes, this carries some risk, but perhaps less risk than driving yourself to inaction!
Particularly if I know that I won't start the project for weeks or months, then I often don't bother writing down anything more than the tangible result of completing the project. I might write down one particularly risky task that requires long lead time, because that determines when I need a reminder to start thinking about it. Let's say that I plan to do a home renovation sometime in the next two years and I need to find a trustworthy lead contractor to act as "technical lead" on the project. Suppose it's June 2014 and I don't plan to start the project until March 2015 at the earliest. I would create a project called "Renovate Top Floor" with a tangible result of "fully-rebuilt top floor with robust solution to leaking problem around the chimney" and a single task of "Choose a lead contractor for the job". I think I need 2 months to choose a lead contractor, and I really don't want to start now, but I also know that the person I find will probably be booked, so I add another 3 months' lead time, just to be safe. This means that I set a weekly reminder starting October 1, 2014 to look for a contractor, then forget about the whole thing. If, by chance, I find myself in a conversation about contractors in August, then it's possible that someone I trust will recommend a contractor that I feel comfortable trusting, but otherwise, I can ignore the issue until October, when I really need to get started. (Getting Things Done calls this reminder a "tickler".)
Most importantly, I don't stop now to try to figure out everything that needs to happen in the renovation. I know the result I want, I know the one action that I probably need to do next, and I write that down, set a reminder so that I start thinking about it in details when I need to, then I move on with my life. Calmly.