Your understanding is correct (mine, in my first post, was wrong). Here's how I think the long-shadow works: You can get an approximate E/W line using any two marks. The closer to solar noon, the closer to reality the E/W line will be. To get an exact E/W line, you'll need one mark some time before solar noon, and a second mark the same amount of time after solar noon.

I think. This is a lot more astronomy than I ever learned in school <img src="/images/graemlins/smile.gif" alt="" />