Overall I think what we need to for.is create such strong identities for our sites that once people interested in a particular topic find us through Google, a link another site, or any other means, they feel compelled to bookmark, subscribe, or take other action to keep coming back.

I finally got around to setting up Facebook and twitter feeds for my sites a few months ago, realizing they're the new RSS. Every day I get a few new followers/fans. It's slow going and it will be months before I top 1000 on either network at this point – the process of gaining my first thousand RSS subscribers in 2008 was much faster. But at least the number is going up instead of down, which means the content still strikes a chord.

Perhaps broader… or narrower content would attract even more people. Watching newer niche sites like Android Police it new broad topic sites like The Verge is intructive — although The Verge is clearly playing with a loaded deck since the all-star writing team brings a lot of existing fans to the table.

Even that isn't worth what it once was, because most sites will summarize your story so thoroughly that few readers will follow the link back to your site (although if you create a youtube video at least they might embed it).
All of which is to say, we're at Google's mercy – or we need to build rabid canvases somehow.

I sometimes regret that my big idea for mobiputing was to launch a platform agnostic blog about mobile apps. It's much easier to get readers to identify with your brand when you're writing about the device or operating system they already use and love.

It's possible if I'd done absolutely nothing and hadn't spent the last few months trying to guess what Google wants, traffic would be even higher right now. 

Still, it's good to see that Google is specifically looking for ways to avoid false positives when identifying low quality sites.