I recall the day I didn't stop thinking how to break a problem down. I kept chipping away till each bit was dead simple. It was absolutely magnificent! It could be applied to everything! But within a month, I'd lost what I took for granted: knowing the dots would connect. (Thank you, wtvanhest; that quote struck a chord.)
I would chop up a big problem into beautifully simple tasks, but partway through executing them, I'd begin to worry about another task. Maybe it wasn't the best solution. Maybe it would be better this way, or if I changed dada-dada... I would take the simple solutions and mangle their limbs, building another big, complex, can't-fit-in-my-head problem, while still retaining the notion I was being effective.
I solved this problem (for me) by keeping a "complex problem" journal. I write roughly what the problem is and why it's a problem. I write how I'd know the problem is solved. Then I break it down into bits, writing a few sentences on how the bit connects to other bits. Sometimes it takes an hour, other times a day or a week. But once it's obvious, when my replies to any question are "yes, the sky is blue" and "it's because fire is hot", I need only reread the journal (often) to keep it fresh.
I accomplish this through delegation. If you leave your keys in a bowl near the door, you're prone to walk right past it, because you need to remember "my keys are in the bowl" every time. And you will remember it, but sometimes you won't. I leave my keys hanging on the door handle, so it's impossible to forget them -- they fall when I open the door. I don't rely on my faulty memory, I rely on my senses and the ability of the universe to trigger them.
It was a wonderful article, and I'm always thankful for whatever tests my assumptions and manner of doing. After all, the nicest thing anyone can do is knock you off your track.