Wow, that's a great summarization of all the persistence technologies available.
I found the stats at the end to be deceptive though. You'd have to code support for 3 different native client storage APIs in JavaScript + use Flash in order to cover enough users to rival cookies. It's no wonder most web devs just use cookies.
The end result isn't an alternative with a 1% gain over cookies. The end result is a system that writes & reads to all possible data stores available simultaneously. The gain from that is a site where an "always logged in" can be potentially much, much longer.
There are still a large number of people who have terrible problems managing web site logins. Reducing the barriers to using your site for returning visitors is definitely a goal with a lot of value and clear ROI.
To my mind, this isn't a problem with just the browsers, but with Flash as well. An in-browser cookie flush shouldn't screw with plugins unless they specifically request it. Thus, I think that there really should be an API for that kind of thing.
That's pretty interesting, I didn't even know about Flash cookies. Feels kind of weird to learn about technology first hand from the NYT, but pretty cool nonetheless.
One heavy handed way around stuff like this is to use a VM (with shared storage to your laptop) for browsing and rolling back at the start of a session. Something polished and stable like VMware on Linux has never been overly annoying for me.
Paul from Meebo gave a great presentation on all the sneaky persistence options a few months ago: http://www.slideshare.net/idontsmoke/client-side-storage
Interesting takeaway: more visitors have some form of persistent storage than have cookies enabled.