Martyn,

Advice from Mike is similar to what I would advise.  Expanding on one point:

>I am assuming (within practical limits) that as much capacitance as
possible in the antenna is beneficial. It will resonate with lower
inductance and therefore lower resistive losses.

Yes, that is correct.

Correct in respect of an ohm or two in coil loss (equivalent series R), but
more complex in terms of net ground loss which is likely in the range of
tens of ohms to a hundred ohms, and is unlikely to change much with
variation in spacing of parallel wire top loading.  The capacitance should
be maximised in the top loading, but not in the "up-wire" (multiple
up-wires, if needed, should taper to a single connection point at the lower
end).  Within reasonable limits, the up-wire should be clear of the house
and tree, and rise vertically from the top of a ground mounted loading coil.
The "environmental soakage" from the lower part of the up-wire (which is all
at high voltage) is another insideous factor that increases equivalent
series resistance.

73, Bob ZL2CA