More design questions

I am trying to understand what value (mfd) the rolled capacitors will be when
I get them done.  I believe you said in earlier correspondence that they
would be about .02mfd.  I calculate about .009 mfd.  The formula should be
C(pf) = (.224KA/d) N-1  where N = number of plates.  For this capacitor:  K =
2  A= 1260 (aluminum 14"x90" for 4x8 sheets of poly)  d = .063  We probably
won't achieve a true distance between plates of .063 (with .0625 material)
because it probably is not practical to expect to get it rolled that tight.
 I would expect the actual value to be somewhat less than the calculated
number of .009mfd.  Have I gone wrong somewhere?

I also need some education on spark gaps.  I have built one cylindrical gap
using 6.0" dia PVC pipe and 7 copper pipe sections of 1.5"dia and 3.0" long.
 I hooked this up as a dead short across a 12kv 60ma transformer and it seems
to work fine.  According to the spark gap instructions, I will need two of
these in series for 12kv at 720VA to 2.0KVA - so I am building another one.
 Some of what I have read tells me that the spark gap is used to help tune
the Tesla coil.  I don't see how varying the spark gap would change the
resonante frequency of the tank circuit but I do see how it would control the
amount of energy introduced into the circuit.  If the gap is too narrow, it
will fire before the capacitor is charged to the full potential available and
the coil will not be as efficient as it could be.  If the gap is too wide I
think it would operate erratically if at all.  If this is true, why do we
want to use a fixed gap system?  With two cylindrical gaps in series set at
.030 each we get a total gap of .36 inches.  Do I really need to add a single
fixed gap in series with all this to "tune" it?

Thanks,  Ed