I used SG MJ data from EDshipyard. Selected a ship, SG, power distributor and shield cell banks. (I picked the Asp because I could install all 5 cell banks A-E at the same time, with only one energized at a time.)
The method was to De-energize and reenergize my SG. Placed 0 pips in SYS to allow SYS capacitor to empty. (As mentioned before there are 3 charging phases, recap: first stage is 16 seconds and nothing charges during this warm up phase, 2nd stage draws the bars in a clockwise ring charging at around 1MJ/s, 3rd stage begins at 1/2 full and charges at normal rate.) After stage one is complete, the sys capacitor will empty in a few seconds allowing me to manually adjust pips to a recharge rate of my choosing (less than 1 MJ/s, so as to not charge up the capacitor.) I removed SYS pips immediately upon reaching the end of stage 2, which halts shield charging at 1/2 full.
I allowed 1 shield cell to Charge shields. At this point I place pips in Sys at a rate close to but less than .6MJ/s (1 pip on 4A pd is equal to .475 MJ/s in accordance with game stats) and measured the time it took to fully recharge. In order to determine fully recharged, the capacitor bank will have charged up one bar (the time it takes to charge one bar was not taken into consideration for this experiment, but can be simply calculated. From 0-1 bar = 1/10 of 23 MJ = 2.3 mj/.475mj/s = ~4.8 seconds.)
So what we are left with is an amount of time and a recharge rate. Multiplying the time (minus the 1 bar charge time) by the recharge rate will give you the amount of MJ that was not charged by the shield cell. Subtract this number from a 1/2 a full shield (since we started at half full) and what you are left with is the amount of MJ charged by the shield cell.
I do need to go back and reconfirm these numbers, but they should be fairly close. And now that shield boosters are out I can test this procedure out on higher ranked shield cells.