Yeah, I figured there must be some leakage that finally cascades into the breakdown and sparks as the voltage rises.

I would not call it "leakage" , because it is strictly a function of voltage. Leakage is quite different, often found in real world situations. Charge transfer is pure physics and can happen in a "perfect" system with zero leakage.
Certainly leakage is a suspect, but actual charge transfer is.
Lightning is a good example of charge build-up until breakdown, and it has been seen to pass through intermediate conductor items and continue to earth.

 

I would not call it "leakage" , because it is strictly a function of voltage. Leakage is quite different, often found in real world situations. Charge transfer is pure physics and can happen in a "perfect" system with zero leakage.
Certainly leakage is a suspect, but actual charge transfer is.
Lightning is a good example of charge build-up until breakdown, and it has been seen to pass through intermediate conductor items and continue to earth.

Voltage flashover even in ultra high vacuum systems is common where there is a discontinuity on the surface from a particle or other surface artifact. The field gradient at the point starts to generate cold Field electron emission.
https://farside.ph.utexas.edu/teaching/qmech/Quantum/node50.html

It's a form of "leakage" current that's normal in high voltage, high vacuum systems. If this emission current increases to the next stage, where excessive Field electron emission starts to heat the surface of the electrode into thermionic emission bad things start to happen. That's usually when we see a exponential increase in 'leakage' leading to a arcing condition sustained by plasma generation on the current generation sites.

 

 

The voltage will build up across the first gap, which will also cause voltage to built up across the second gap, due to an initial, small ionizing current through the gaps as it nears the breakdown voltage of the gaps.
They will then fire together when the voltage is equal to the gap breakdown voltage of both added together.

Obviously it's not possible for the gaps never to fire, just because they are in series.

Great! Now one more thing,- It just seems to me, that If the positive high volts fires first gap as wide, and 2nd gap to ground is narrow, and in series, doesn't the final spark gap fire at same high voltage as first? The first gap has to fire to GET to the second gap to ground, and both points (gaps) are electrically connected together? In other words, final gap has to be smaller, but with higher voltage as fired. I forgot to mention that in the old drawing circuit, there was a 1.75 millihenry inductor in between the two spark gaps. Maybe the back EMF going forward also, added a higher, high voltage ignition. Of course a CDI unit can fire spark plug with higher voltage at plug with narrow gap, but that is always in microseconds, not regular needed milliseconds.

 

It just seems to me, that If the positive high volts fires first gap as wide

Why do you say that gap fires first?
They both fire at the same time (or at least within the time it takes a speed-of-light wave to travel the distance between the two gaps).

 

"" Of course a CDI unit can fire spark plug with higher voltage at plug with narrow gap,
but that is always in microseconds, not regular needed milliseconds. ""

The Spark-Plug-Gap, and the density of the Air/Fuel-Mixture in the Gap,
are the only determining factors of the Voltage across the Spark-Plug-Gap.
The Ignition-System generating the Electromotive-Force does NOT determine the
Voltage across the Spark-Plug-Gap.
The only requirement of the Ignition system is that it has
the capability of generating enough Voltage to jump the Spark-Plug-Gap,
under the existing conditions in the Combustion-Chamber,
and these conditions change radically during various normal driving conditions,
causing the required Voltage across the Spark-Plugs to change substantially.
------------------------------------------------------------------------------------------------------

At high Engine RPMs, multiple "Milliseconds" of time are not available,
so a "longer" spark duration is of very limited advantage.
Generally, a longer spark duration would only be advantageous in an
Engine with a poorly designed Combustion-Chamber,
or, in a situation where the Engine is in a very poor state of tune.
.
.
.

 

Really, a longer spark duration was sometimes a benefit in carburetor equipped engines in which the starting sequence required pumping air through the cylinder prior to the fuel-air mix reaching it. Consider some of those antique vehicles where the engine had actual priming cups on the intake manifold. The other claimed possible benefit of the dual gap arrangement was the claimed faster rise time of the spark, possibly helpful with fouled spark plugs. BUT for engines that tended to foul the plugs with conductive carbon, A good CD system would deliver a lot more power, and that was a much better solution.

 

The voltage will build up across the first gap, which will also cause voltage to built up across the second gap, due to an initial, small ionizing current through the gaps as it nears the breakdown voltage of the gaps.
They will then fire together when the voltage is equal to the gap breakdown voltage of both added together.

Obviously it's not possible for the gaps never to fire, just because they are in series.

It just seems to me, that If the positive high volts fires first gap as wide, and 2nd gap to ground is narrow, and in series, doesn't the final spark gap fire at same high voltage as first? The first gap has to fire to GET to the second gap to ground, and both points (gaps) are electrically connected together? In other words, final gap has to be smaller, but with higher voltage as fired. I forgot to mention that in the old drawing circuit, there was a 1.75 millihenry inductor in between the two spark gaps. Maybe the back EMF going forward also, added a higher, high voltage ignition. Of course a CDI unit can fire spark plug with higher voltage at plug with narrow gap, but that is always in microseconds, not regular needed milliseconds.

 

What problem are You trying to solve ???? ( with details please ).
.
.
.