Michael,

This is an area of interest ot me as well.  Generally, the more lements you
have, the better as far as gain and a tighter pattern.  However, in
designing multiband yagis, there are always compromises.  If you have
trapped elements, the traps introduce loss and compromises in terms of
element spacing.

Generally, the better-performing antennas are on longer booms.  However, a
consequential factor to consider is weight, and more importantly, wind
loading.  Longer booms and elements result in high surface areas, which
create wind load and torquing problems for rotators and antenna/tower design
safety.

In terms of what beam to get, then, you have to weigh all these
considerations.  Also, in looking at specifications, make sure you are
comparing the same measurements.  Some manufacturers inflate their gain
figures by quoting free-space gain over an isotropic source, which in my
opinion is'nt real-life.  And, others quote gain over a dipole.  The gain
over a dipole at 60 feet, for example, will be considerably less than
isotropic gain sinc ethe dipole will have an ingerent gain of 3DB.  So, make
sure when you look at gain quotes, that you are looking at the same figures.
Force 12, which are reputedly the best yagis made (at least judging from
what many DXpeditions carry with them), quote their gain in terms of a
reference dipole which makes their gain figures look a lot less than a
HyGain or Mosley.  Yet, if you look at independent comparisons such as Steve
Morrison does, they perform much better.

A second and more important considreation is antenna pattern.  This is more
critical than a DB of forward gain here or there, in my opinion.  Here,
again, some manufacturers inflate their figures by citing the best-case
front-to-back ratio or they may design their beam to have a deep null at the
back, but the secondary lobes off the side aren't really much lower.  So,
the pattern is poor.