@VikeMan
I have always given for granted that it's a glucose not sucrose solution that the refractometer is calibrated for. But I also guess there is little difference between the two. I'll have to check my refractometer's instruction.
What I mean is that the SG and the Brix values are equally exact, or equally wrong.
If you apply a "Brix correction factor" you arrive to a more exact measure of density, but then again wouldn't be more exact than a "SG correction factor" which would make you arrive to the same density.
The fact is, the refractometers that we buy are mostly used for wine work where Brix dominates and are normally calibrated in Brix. The calculator is therefore made for Brix calculation.
In this sense, I understand what you mean, that ultimately you don't use the SG scale, because you will have to correct the value with the Brix calculator.
My observation was due to reading often that the SG measure is less precise and one should only use the Brix value of the refractometer, because the Brix value is "native" and the SG value is approximate.
I now see that this is not what you were saying.
If we take this calculator:
Refractometer Calculator - Brewer's Friend
And we use at OG the measure taken with refractometer in SG it's equally precise (or imprecise because the wort is different from the sugar solution) than the density taken in Brix with the same instrument. But the final density, to be used with this calculator, must be taken in Brix, because the calculator doesn't give us the possibility to use SG as the final density.
The "correction factor" is actually quite a misnomer. Sometimes I find it should be less than 1, sometimes more than 1 (when comparing OG taken with refractometer and hydrometer). I tend to ignore it. I accept the datum from the refractometer is approximated and I don't know in which direction. But yes, if you want to apply a "correction factor" and you want to use this calculator you have to work in Brix.