That is a great question. The “simple” answer is that no, hydrometry and refractometry are not interchangeable and that you shouldn’t try to use a refractometer during active fermentation. Refractometry relies on measuring how a ray of light will bend (refract) through liquids of different densities to be read out on a scale. Hydrometry relies directly on density of a solution itself and accounts for a density change as the sugar disappears and alcohol (alcohol is less dense than water) is generated.
When I was a winemaking student at UC-Davis we were taught that refractometers are best used for unfermented juice as an initial Brix reading and that once fermentation started, hydrometers were the best tool to monitor fermentations. This is because alcohol and carbon dioxide bubbles will both interfere with how the ray of light passes through the refractometer and are difficult to control for. However, as you bring up, there should be a way to calculate the effect of the alcohol to arrive at a “corrected” sugar level.
Some people have attempted to do so. There is an interesting post I found on www.valleyvintner.com, the website of The Valley Vintner, a respected winemaking supply shop based in Livermore, California. They did just what you are proposing and used a refractometer alongside a hydrometer on a test fermentation. Then they tried to “calculate out” the increasing alcohol level (and decreasing density) to create a relative Brix (or specific gravity) using a refractometer. The math is complicated but to make it a little simpler the good folks at The Valley Vintner created a spreadsheet that you can use to correlate a refractometer reading to the true sugar level. You can find that here: http://valleyvintner.com/Refrac_Hydro/Refract_Hydro.htm
In all of the wineries where I work, we either use hydrometers or a digital densitometer like the Anton Paar DM35 handheld model to monitor fermentations. Hydrometers, albeit a little old fashioned, seem to work well to monitor the disappearance of sugar over time and are relatively inexpensive. They are, however, very breakable and the readings have to be corrected for any temperature variation from 68 °F (20 °C). The Anton Paar gadget is a couple thousand dollars but if you’re doing hundreds of samples during the morning “Brix run” for weeks on end during harvest, they are worth their weight in gold.
So, should you try to use a refractometer to monitor your fermentations? It’s up to you. You could be like the folks at The Valley Vintner and give it a try to see how well it correlates with your hydrometer measurements. The carbon dioxide bubbles are hard to correct for on a refractometer and the Valley Vintner’s method doesn’t seem to try to correct for them. If you are using a hydrometer, be aware that high solids in your juice sample, as well as carbon dioxide bubbles that might adhere to the hydrometer stem may cause it to read “lighter” (float higher) than the sample actually is. It just seems like no method is perfect.
I will say this, however. In my opinion, it’s OK if Brix during fermentation is not a 100% accurate number. The most important thing you are monitoring is the progress of the fermentation, not necessarily the absolute sugar number. That may sound a little strange, but for example, I’m much more concerned over whether my fermentation is active and my Brix has dropped 1.7 degrees in 24 hours rather than is it exactly 19.4 Brix right at this moment. For me, it’s about monitoring the health and completion of a fermentation rather than an exact number at any given time. Once your hydrometer readings bottom out at -1.5 for three days in a row you know you’re probably pretty “dry” and it’s time to start using Clinitest tablets to test for minute amounts of reducing sugar anyway. In commercial wineries, when we get to this point, we send a sample of the new wine to the lab for a Glucose + Fructose test to see how dry we really are. Most wineries consider anything under 0.15% (1.5 g/L) sugar to be “dry.”