The Hanna Ca and Mg checkers haven't been terribly well received by the community. Many (most?) advise against them, and I believe a big part of the concern is the ease at which measurement error can be introduced -- compared to the more standard titration testing.
These are easy titration tests to do and difficult to mess up, you're just looking for a (pretty noticeable) color change after adding reagent drop-by-drop. You count the drops it took to observe the color change, then reference a look-up table to convert # of drops to measurement parameter.
Bigger picture -- It's important to develop a testing process that is repeatable for you. For instance, some people use a syringe to measure tank water. Others use the cuvette lines. Either one can work -- as long as you do it the same way every single time. My method is to use the cuvette line, and fill until the bottom of the meniscus is level with the cuvette line.
I also wipe down the cuvette exterior before the standardization and also before the measurement. Additionally, I use the same cuvette each time I test -- I marked the bottom of 1 ea cuvette with "A", "N", and "P" with a paint marker. And I place the cuvette in the checker such that the "10 mL" text is centered, facing me. Why? Because manufacturing tolerances mean there are slight differences between each cuvette -- even at different parts of the same cuvette. And fingerprints/grease are known to impact results. If your first measurement (C1, used for standardization) is performed with a fingerprint in the measurement path, and your 2nd measurement (C2, the actual measurement) is taken without the fingerprint -- you're going to get erroneous data.
I also rinse cuvettes immediately after testing (to prevent tinting) and store filled with RODI water (which is great at dissolving stuff).
Regarding salinity measurements with a refractometer -- I verify calibration with reference solution before every measurement "session." I also verify after if the results don't match what I expect. I allow the sample to "sit" on the refractometer for a minute or so before testing. I use the same light source for every measurement, and I wipe it clean and rinse with RODI prior to storage. I've seen refractometers drift over time, and this process is simple enough and eliminates uncertainty.
I do not deviate from these processes, ever. And my results are very repeatable.