There's a correction in the August 2009 QST to his article in the July issue.
See "Feedback" in QST or on the ARRL web page:http://remote.arrl.org/qst/feedback/2009/08/
"In “A High Gain Single Wire Beam” [Jul 2009, pp 38-39], the author notes that the inductor should be in series with the center conductor of the coax, not across the coax. The resonant frequency, that at which the SWR is lowest, may need to be adjusted by adding or subtracting turns from the coil. In his case, on 20 meters, the coil only needed about 8 turns rather than 26 turns. If the inductor is in parallel, the antenna will work, but the SWR will likely be about 3 to 1."
The original looked wrong to me too.