It's puzzling how Google works sometimes.
Yep... That's the 'POLITE' way of putting it...
It just highlights how many loopholes there are in G's algo. But trying to mimic them or repeat them is impossible: I've done a number of case-studies and tests with this. But you can never replicate anywhere near enough of the factors - they're too diverse.
It's luck of the draw sometimes, and one of the reasons why when I go after a certain niche/keyword set, I do multiple sites:
1) As insurance,
2) as some will do better than others randomly,
3) as I want multiple sites on page 1.
I ALWAYS go for 3+ different domains in top 10 on a keyword I want. That way, I know I'll get decent traffic.
BTW: There've been some interesting releases on traffic stats that I've seen. Even though the no.1 position for traffic still shows about 42%, there's a lot of evidence to show how people's search patterns have evolved over the last 2 years, and most now click on at least 3 links. And the 42% stat doesn't account for that.
This means that you might not be the first click, but if you've got several in the top 10, you're almost guaranteed to get the majority of searchers for that term at some point. And of course, if you're better/more insightful and more engaging than the others...
It's another argument for quality content.
I think it's more important than ever to have good readable content - even if it's spun (which it mostly will be, let's be honest). So good spinning technology and a strong initial content quality is paramount.
...Just my 2 cents...
