I do not blame anyone for having an incomplete understanding of physics. No person on earth has a complete understanding of physics. What is frustrating is listening to people talk about physics without doing any background work. It is detrimental to mankind to speak officiously when you really do not know a subject.
Some steering systems operate entirely upon energy harnessed in the form of aerodynamic drag imposed upon a vane. These are usually impractically large, barely effective, aesthetically displeasing, and only work on very well balanced boats.
Most steering systems are servo-pendulum type, with a windvane, a servo-oar, and a connection to the tiller. These work in a clever way. The vane sends energy to the oar. The oar then collects energy from the apparent flow of water caused by the boat making headway. This energy is sent to the tiller through
lines guided by blocks.
There are also other varieties of self steering vanes, the most elegant of which is (imho) the trim tab on an outboard rudder.
This is a great resource when learning the basics of wind vane design.
http://www.selfsteer.com/windvanes101/index.php
Does anyone have experience with fenestrated rudders? Reportedly they require less input force from the helmsman, but have the same steering effect as a normal rudder. If this data is accurate, then they could be of great use in self steering systems.