Here is my take on vertical tail area.

Free flight models often have very small vertical tails and lots of polyhedral. The generous polyhedral keeps them from spiralling in when the plane is banked by lift on one side of the wing. The small vertical tail puts them on the verge of yaw instability (dutch roll) but, allows the model to follow lift rather than just boring through it.

R/C thermal soarers need larger vertical tails to give them the necessary yaw control authority without very large, drag producing, rudder deflections. Most R/C thermal soarers would benefit from vertical tails of higher aspect ratio so that the induced drag oassociated with rudder deflection was less and the yaw control authority could be achieved with less vertical tail area. Generous polyhedral produces stronger yaw to roll coupling. This reduces the necessary yaw angles and reduces the drag associated with a yawed fuselage and with a highly deflected rudder. The extra parasitic drag of a large vertical tail has to be ballanced against the drag reduction associated with maneuvering. Where this balance should fall depends on how much time the plane spends maneuvering versus the total flight time. Not an easy question to find a neat answer to.

You can learn a lot by studying the designs of Dr. Mark Drela and reading the Ask Joe and Don section of the DJ Aerotech web site. See:
http://www.charlesriverrc.org/articles.htm
http://www.djaerotech.com/dj_askjd/askdesign.html
http://www.polecataero.com/ (go to articles)
http://www.monkeytumble.com/hlg/supergee.htm

Dr. Drela's designs are the most thoroughly engineered, refined, structurally superior and aerodynamically superior that I have seen in over 35 years of R/C modeling. I don't believe I am overstating my judgement. There is a treasure house of information in his designs.