My expectation is its not better than an eyeball guess.
It depends on how precisely you can measure salinity, but let's look.
Suppose the tank volume is actually 100 gallons and is exactly 35 ppt (sg = 1.0264)
If you add 2 gallons of RO/DI, the salinity becomes 34.31 and the sg is 1.02588.
That change is a sg difference of 0.00052
How well can you reliably measure salinity?
Suppose it is very good and can distinguish 1.0250 from 1.0251. That difference of 0.0001 is 1/5th of the total. So the relative error is about 20%. 100 gallons +/- 20 gallons.
I think that is a correct analysis, but if anyone sees a problem, be sure to note it here.