Following up on this, I’ve been told my approach was indeed correct, but apparently the public CAD model has the blades slightly modified for IP reasons. The measurement from the actual design file has been confirmed as 14°, which I’ve now added to the technical details on the product page 
This is effectively a re-arrangement of my approach in this T200 comment, but with an included efficiency estimate (so you’re not the only one doing such calculations)
I’m assuming you meant \arctan here (i.e. \tan^{-1}), not hyperbolic tangent (\tanh)? That matches your calculated value, and makes follows your triangle logic description.
There’s a formatting section in the How to Use the Forums post, which covers the basics - the maths is essentially handled by LaTeX syntax, if you’re familiar with that