YIKES! Screws in the fittings!

Having carried with my antennas about a fair bit, I rely on
the slip fit to hold the antenna together just fine in the
field, while still allowing me to pull it apart when I
need to stuff it in my carry-on luggage.
Using standard PVC pipe fittings, a 1" tape measure
element will fit inside of a 3/4" fitting, and a 5/8" tape
will fit inside a 1/2" fitting. This allows me to fold
the elements in thirds, rather than in half, so it stores
in an even smaller space. See the 6th photo here:
http://members.aol.com/homingin/ncohpix2.htmlRegarding different dimensions: the WB2HOL design is
specifically for a good null off the back, while other
dimensions may be optimized for gain, 3dB beamwidth,
or lowest side lobes (which will reduce the chance of
taking a bearing on a spurious lobe.) Personally, I
never use the rear null for ARDF, so my criteria are
"clean pattern" (all spurious lobes down 20dB or more)
and reasonable gain consistant with boom length. One
problem I sometimes have with the WB2HOL dimensions is
finding a place to mount the receiver and compass while
still having room to hold the boom near the center of
gravity. Using a wider spacing between the driven
element and the reflector helps this. (In the photo
you can see that I put the compass near the driven
element - but not too close, since the element is steel -
and the receiver on a short piece of pipe behind the
reflector. This specific antenna uses equal spacing,
and is developed from a W4RNL design.)
I've seen several European designs (which probably
pre-date the availability of yagi modelling software)
where the driven element is much closer to the director
than the reflector. In fact there are many different
yagi designs used in competition, from 2 elements to
5 or 6. Part of the choice will depend on how well your
equipment can indicate a 1dB signal strength difference:
if this is difficult, then the half-power beamwidth is
critical. The VK3YNG and VK4BRG receivers with a VCO
make this relatively easy, so the sharpness of the main
lobe isn't as critical for bearing accuracy.
It is all a bunch of trade-offs. There is no "right
answer" or "best design"!