Not so. Time to put this misconception to rest.
Read my description of K9AY Loop operation in the most recent ARRL Antenna
Book and review the description of a DF loop with sense antenna in that,
or earlier antenna books.
I have never been able to achieve the same pattern as the loop using
close-spaced elements in a phased array. However, a model using a small
loop in combination with a nearby short vertical does indeed result in the
same pattern. (Feed them with equal currents and 90 deg. phase shift.)
Magnetic lines of force from an arriving wavefront MUST induce currents in
a loop. You can't ignore them and assume that the vertical end portions
are the only place energy is captured.
I certainly am not ignoring the horizontal wires. The horizontal component
of wires, besides supplying phase delay and phase inversion (acting like
leaky wide-spaced feedlines) adds significant higher angle component to the
response, which is generally not what we are looking for.
The meaningful lower-angle response all comes from the vertical tilt of
As a matter of fact the more perfect the earth below the antenna, the more
suppressed the high angle content.
This can be clearly seen by anyone if they model a vertical antenna, and
they models a vertical radius loop antenna. The vertical has a null straight
up. The loop has a strong response straight up. I can't imagine the
straight-up response added by the horizontal component of the wire
contributes much to DX response.