I've always used RO water to calibrate or check my refractometer. Could never bring myself to bother ordering & trying calibration fluid despite reading all the dire warnings & horror stories on various forums. I've also used a floating hydrometer as well as the swing arms. Tank has been running for 4-5 years with no apparent issues related to out of whack salinity.
The two swing arms I have both yield different readings with the same test water. I don't like them, they're crap, even though I put correction marks on them using the refractometer reading as a calibration standard. In a pinch I'll use one & go by the mark I put on them, although I'm not sure why I haven't thrown them in the trash.
I broke the floating hydrometer a while back, but actually really liked it. Old tech but very accurate. The only thing you need to watch for is solution temperature since the floaters have no temperature compensation like most refractometers do. The floaters are designed, calibrated & tested at a specific temperature & the instructions usually include a temperature compensation chart. Provided the solution you're testing is around 20C, the floating hydrometer will be bang on. I use one regularly for my wine & beer making efforts, I think I recall using it on the tank water & it was accurate.
Ah yes, the refractometer. I have the instructions for mine & the calibration procedure calls for distilled water, no mention of any other claibration fluid. RO or RODI water should be a fair substitute I would think. My refractometer is temperature compensated, so provided I leave the solution on the prism glass for at least 30 seconds to adjust to the ambient temperature of the refractometer, the reading will be accurate. I've read the arguments about refractometer scales not being linear, but from my own experience, if I have a reading one notch above zero with RO water, then test tank water at 1.026, if I tweak the refracto to read zero with RO water, a subsequent test of tank water gives me 1.025. Quite linear I would say. No need to waste time & $$$s on calibration fluid that checks your refractometer at 1.026 or whatever.
Here are the calibration instructions for my refractometer verbatim:
1) Open daylight plate, and place 2-3 drops of distilled water on the main prism. Close the daylight plate so the water spreads across the entire surface of the prism without air bubbles or dry spots. Allow the sample to rest on the the prism for 30 seconds before going to step #2. (This allows the sample to adjust to the ambient temperature of the refractometer).
2) Hold daylight plate in the direction of a light source and look into the eyepiece. You will see a circular field with graduations down the center (you may have to focus the eyepiece to clearly see the graduations). The upper portion of the field should be blue, while the lower portion should be white.
3) Because this instrument is equipped with Automatic Temperature Compensation, the ambient working temperature of the room must be 20C (68F) whenever the instrument is recalibrated. Once calibrated, shifts in ambient temperature within the acceptable range, should not affect accuracy (10-30C). Using distilled water as a sample, look into the eyepiece and turn the Calibration Screw until the boundary between the upper blue field and the lower white field meet exactly on the 0.0 ppt (or 1.000 spec. grav.).
So it appears quite important that the calibration be carried out at the proper ambient temperature 20C. No mention of calibration fluid whatsoever.
Photos of readings years ago with the two swing arms & the floating hydrometer using the same tank water. Don't recall what the refractometer read exactly, but do remember that the floating hydrometer read virtually the same.
Coralife swingarm
IO Deep Six swing arm
Made in Taiwan floating hydrometer
IIRC, the refractometer read right around 1.025 with the same water. Result... Coralife low by 2-3 points, IO a tad more than 1 point low & floater just a fraction under refracto reading. Unlikely that I'll be ordering any calibration fluid.