I recently did this on my own tank but did it with a reverse mentality. Probably a bit more complex but Instead of adding chemical I measure changes in salinity against top offs. My tank is 328 gallons and I run about another 68 through my sump (hold about 105) . This gives me a potential max volume of 396 that i knew it could not exceed. Now I have about 300 to 400 pounds of live rock in it as well. Since my tank goes through about 5 gallons a day of evaporation, I used a red sea blue bucket of salt which happens to be exactly 5 gallons. For my testing, I filled it every day till it overflowed and then scooped out 1 gallon ish into a smaller pail just so I can transfer it to tank without spelling, but ensure I was still adding exactly 5 gallons consistently
To estimate my volume, I basically use a tropic Marin hydrometer and a temperature probe to take sg level daily. Since it hard to get exact level i took the SG range that the water level was between ie. 1.0251-1.0252…. After that took the current temp. Using and old chart Randy Holmes provided in an article I map that into excel for the temp ranges only that I used and use linear trend (not 100 accurate but delta are so small between values works for estimates) to break it down to .1 decimal ranges for temp and .001 ranges for SG. After ward I used some excel formulas to map my SG and Temps reading to PPT.
After topping off my tan, waited like two hours till my apex salinity probe stabilized and took another temp and reading with the TM. Afte my spread sheet calculated the PPTs of the second reading. I added a formula to calculate the percent changes of parts with respect to knowing that the change was due to 5 additional gallons of RO water was added did it 4 times using high range value test b4 to high Value after , low to low , high to low and low to high (won’t get into the Math unless asked).
After I had 4 different water estimates I then discarded every value I know could never exceed my true tank water only and sump only volume (up to current sump water only). And average the rest.
After about three weeks of test seem to average right around 350 Gallons (P1) on my spread sheet which was roughly around what I was expecting and have been rolling with since.
Dosing does seem to match product expectations when accounting for this value now.