Well I stand corrected. Assumed that the alcohol & yeasties would interfere with the refractometer too much. Found this reference which explains things pretty nicely:
Paper No (misco.com)
This document is a nice selling point for Misco for their refractometer, but it certainly shows how useful are refractometers for vineyard work.
Beer wort is a bit more complicated, because we don't deal with a sugar-water mixture, but a complex sugar - simple sugar - protein - whateverelse mixture, and the refractive index of the latter is different from the refractive index of the former.
In order to resolve the difficulty, it is customary to apply a "correction index" to the measurement that the refractomer gives. That means that there are two calculations to be made to infer the actual density of a beer from a refractometer reading:
The first correction is the "correction index", which is normally 1,04, but which is dependent from the single wort and this means that there is a degree of approximation which is not eliminable;
The second correction is the equation which is used to derive the density from the observed and "corrected" value given by the refractomer. This equation tries to keep into account the alcohol content inside the sample. Various equations have been proposed: Novotny's equation is the "latest fashion".
Winemaker have an easier life with refractometers, but they make wine only once a year and are less concerned about wasted wort than ourselves