I could recommend to see recent publications by Jon Kjellin, Anders Goude and Fredrik Bülow at Uppsala university (for a complete list from the Division of Electricity at Uppsala university, please see www.el.angstrom.uu.se/publications). We are a relatively large group working on VAWTs (although I'm in wave energy). For a small-scale (12 kW) straight bladed directly driven lift-type VAWT, a Cp of 29% (if I recall correctly) has been verified by experiments (see also www.el.angstrom.uu.se). Later, e.g. Fredrik Bülow has published on a 50 m high 200 kW VAWT erected in South-West of Sweden (in the town of Falkenberg). Please note that there is a very large difference in Cp between drag-type Savonius and lift-typ Darreus and also between Darreus and the so-called H-rotors ("Straight-bladed Darreus"). 11-14% is perhaps valid for drag-type Savonius, but is probably below 10 in many cases. Largest drawback with drag-type VAWT (Savonius) is the high solidity (material use per swept area).
Now, it seems that the wind energy efficiency is about 29% for drag-type VAWT from our simulated computation. Please see: M. Zheng, L.Guo, Y. Li, Y. Tian, H. P. Teng, J. Hu, Y. Zhao, L. J. Yu, Power Efficiency of 5 Blade Drag Type Vertical Axis Wind Turbine,