I was uncomfortable with the variability in my
measurements when measuring the capacitance
of plates with a paper separator, dry and
wet, following the recent thread discussion.
So I repeated the test, with stiff metal plates,
quartz standoffs and distilled water/air as the
dielectric pair.
I was disappointed to find that the capacitance
multiplier for water was around 60, though I
was using a resonance-frequency sensing LC meter
which should have largely neutralized the effects
of parallel resistance from dissociation effects
and impurities in the water.
The moral is, I suppose, "you can't go back..."
but I will look at some other factors in a while.