This is addressed in the article. It says that google uses toolbar latency to determine page load latency and that is the only way toolbar data affects page ranking.
It is widely assumed in SEO circles that Google uses toolbar data, among other sources, for finding new URLs to crawl. They're very enthusiastic about getting all pages on the public Internet into the crawl set. User data gets them there faster and more reliably than a hypothetical competing crawler using only e.g. the observed link graph.
Edit: I did some digging to see if I could find an authoritative source on this, and found that Matt Cutts specifically denies this particular usage of user data for expanding the crawl set on his blog. Mea maxima culpa.
Edit the second: An amusing note on this general subject: Google will fuzz test certain search forms on, e.g., high value government websites to get at the juicy data behind them which would not otherwise be reachable from just traversing the link graph.
I seriously doubt that is the only thing they use the information for, and unless someone can find somewhere in the google toolbar/chrome EULAs that specifically asserts what they will/won't do with the data, I'd assume the they use it for all sorts of things.
I don't think that's true - at least, I've not seen any evidence. There's a good chance the Google toolbar sends back browser data, but I think it's extraordinarily unlikely that Google snoops on what people are searching for on Bing, and what results they end up clicking on, and then adjusts their search results on the back of that.
If you enable the "show PR" thing on it, but that is (or at least was when I last used it) off by default. And Google specifically say they won't use the data for this sort of thing.
