(Didnt completely read the thread, only scanned some posts)
I noticed a couple posts later on listing such references - but a note to be made that I didnt see mentioned. Yes, it is important to remember that temperature is a factor in the calibration of any measuring device for salinity. As temperature changes, the water sample becomes more or less dense as particles more closer or farther apart due to their kinetic energy. A very cold water sample may give a false high salinity reading, where the reverse would be opposit. It was however mentioned that most devices DO have a temperature tolerance range. This is true, but when dealing with different devices for measuring salinity, it only affects certain items adversely.
When calibrating probes, the probe itself is dipped directly into a relatively large sample of water... both are of different temperatures, where the probe cannot change the temperature of the calibration fluid. Hydrometers hold 'relatively' large volumes of fluid, and are made entirely of plastic, thus hold little energy, and will also affect the fluid temperature almost undetectably. With refractometers however, the sample of water found on the glass is so small, and the material the refractometer is made with can hold so much energy, even at 'room temperature' or close, it would change the temperature and therefore density of the water sample or calibration fluid extremely quickly. Even if the temperature of the sample or calibration fluid is more than 20 degrees (F) from the point being measured, the temperature of that fluid upon contact with the glass/plastic/metal will change very quickly, thus temperature is not as much a concern with refractometers, although it is still reasonable to consider it a possible factor in lightly varied measurements.
Hopefully provides a little more insight on the topic.
Cheers,
Chris