As an EE, this is interesting. But I'm not sure about your statement, "if discharged over 1 second " . After the voltage reaches the breakdown value and the arc occurs, I'm not sure how to calculate the time to discharge, but I would think it was less than 1 second. If it is, that is higher instantaneous power (power=watts= joules/second), but if it is a few seconds between discharges, that would lower the average power. So, it depends how fast you can repeat the process. As an example, if it discharged in .1seconds, and recharged/discharged in 3 seconds, the duty cycle would be (0.1)/3 or 0.033. The average power would be .033 of the instantaneous power. I haven't done a detailed analysis, but just offering this for further investigation.

As an EE, this is interesting. But I'm not sure about your statement, "if discharged over 1 second " . I would have to see what the time constant (R*C) is, based on air resistance of the gap, but I would think it was less than 1 second. If it is, that is higher instantaneous power (power=watts= joules/second), but if it is a few seconds between discharges, that would lower the average power. So, it depends how fast you can repeat the process. As an example, if it discharged in .1seconds, and recharged/discharged in 3 seconds, the duty cycle would be (0.1)/3 or 0.033. The average power would be .033 of the instantaneous power. I haven't done a detailed analysis, but just offering this for further investigation.