Jay...In my experience the mathematical adjustment is very close and is meaningful...I have used it many times in test development. As best as I can tell it is linear in behavior at lest in the ranges I have worked with.I should add, I used the caveat of "I think...." in regards to adjusting the salinities of the samples before making comparisons. Is that a valid conclusion? Then, can the adjustment be made mathematically, or would you need to start with samples that were all diluted to the same salinity? Math is not my forte (grin).
Jay