Since you system will not allow you to catch the fuel you flow do it this way.
Weigh the container you put fuel in, weigh the container with fuel in it, the difference is the fuel weight. weigh the leftover fuel, the difference is fuel you put in the cylinder.
Test each side, one at a time... the cloud is N20, you won't see the fuel if it is atomized correctly.
start with the lines full if possible, you want to be as accurate as possible, your pistons are counting on you
Rich is your friend, lean is the enemy, 5:1 has always been my goal for a long button push.
I collected data on every jet, I rechecked my math from my data at every jet change, power went up, ratio stayed the same within a .1 or .2. My last N20 engine failure was a Carrillo rod... bent under the combustion pressure until it hit the crank counterweight.
We killed lots of parts before we stopped believing the jet selection choices from the manufacturers, did the research, built in our own safeties, and tested, and tested, and tested. Some engines might live through the couple of seconds of button that they get drag racing... if you push the button for 10 to 58 seconds, you better have everything right. The fuel system must be fully sorted without the N20, then the N20 and enrichment fuel must be treated as a separate unit, you cannot look at final AFR and say you are good. Don't forget timing, personally I liked to pull 3-5 degrees for N20 in general and then 1 more for every pound per minute that I sprayed. I really can't offer much more than my own experience with our zx-7 and zx-10 motors, other than I used the same ratios with both.