Your proposed method (adding the masses, but using the volume of the solvent) can work under certain circumstances; it is often a good first approximation for the times when it is not a good assumption.
For low solubility salts, you can safely assume that the volume of the solution is the same as the volume of the solvent. "Low" here is somewhat open for debate. As a general rule, I feel safer with this assumption for each order of magnitude below 1 molar / molal. So, I consider this method valid to one (or fewer) sigfigs for solutions near 1.0 molar; 0.1 M makes me feel "2 sig figs safe", etc. (Again, these are just guidelines, don't get hung up on specific numbers.) With a salt that can only dissolve fewer than 10 grams into 100g of water, this method is usually fine to 2 or 3 sig figs. A salt that dissolves less than 1.0g into 100g of water is now getting into the territory of measurement errors of accurately measuring the volume of the solution (in most high school or even some college labs).
Seawater runs about 3% salt to water by mass. Estimating it's density at 1.03 g/mL is pretty valid. In the example in the OP of 80g of salt in 100mL of water (giving a value of 1.8 g/mL) is a good first order approximation, but I would take that value with... a grain of salt.