Essentially, you are correct. In most horizontal phased array systems I
have looked at, the 4.5 dB improvement is a bit optimistic. Some a bit
over 4 dB is more like it, and you can get similar performance from a
driven element-reflector (assuming full size elements) without need for
dealing with the phasing system with a decent operating bandwidth. A
spacing of about 1/8 wl will produce a bit more gain but with the
feedpoint impedance lowered to about 1/2 the single element alone. A
spacing of about 0.15 wl will yield a feedpoint impedance close to that of
the original antenna or only a bit lower, but with slightly reduced gain.
If you want to try the phasing system, you can independently create the
required phase difference with a lumped constant network on the "ground"
and use multiples of 1/2 wl feedlines to transfer the phase settings to
the wire terminals. Or you might try the HB9CV system, which is
well-liked and much used in Europe in both wire and tubing versions.
Whichever way you go, the gain and f-b of the second wire will prove
useful to you.
Incidentally, if you initially opt for the simplicity of parasitical
operation, you can always convert to a phased system later, if you get the
urge to squeeze the last milli-dB of gain from the system.
FAQ on WWW: http://www.contesting.com/towertalkfaq.html
Administrative requests: towertalk-REQUEST@contesting.com