I posted this last night, but it appears to have vanished. It's never as good the second time around...
I generated a fin surface area equation based on depth and length measurements. I used a 2016 Radar fin since I had one laying around and it seemed like a good industry standard shape. The equation obviously only holds for this fin shape, but it should hold good for the trends for any similarly shaped fin. I digitized the fin and used CAD to make the length and depth adjustments. The fin holes are accounted for. The equation is based on a simple linear regression model, which fit the measured data extremely well.
Surface Area = (2.697 * Length) + (4.841 * Depth) - 18.924
The equation fits the measured data with an R-Squared value of 0.999.
Here's the data out of CAD:
Here's the same data in graphical form along with the predicted values from the equation. The measured values are the points, the predicted values are the lines.