There is numerical evidence that
$$\int_0^\pi\arcsin\left(\frac14\sqrt{8-2\sqrt{10-2\sqrt{17-8\cos x}}}\right)dx=\frac{\pi^2}{5}.$$
How can this be proved?
Context
In another question, three random points on a circle are chosen, and tangents to the circle at those points form a triangle, with side lengths $d,e,f$.
The relevant question here is to prove that $P(de<f^2)=\frac35$.
Assuming the radius is $1$, we have
$d=\tan X-\tan(X+Y)$
$e=\tan Y-\tan(X+Y)$
$f=\tan X+\tan Y$
$P(de<f^2)=P((\tan X-\tan(X+Y))(\tan Y-\tan(X+Y))<(\tan X+\tan Y)^2)$
I rotated the graph $45^\circ$ by letting $X=y-x$ and $Y=y+x$.
Then Wolfram gave me $y$ as a function of $x$. After some minor simplifications, I ended up with the integral in this question.
My attempt
Approachzero does not turn up anything similar. Wolfram does not evaluate the anti-derivative.
Here is the graph of $y=\arcsin\left(\frac14\sqrt{8-2\sqrt{10-2\sqrt{17-8\cos x}}}\right)$.
I tried applying some general advice about related integrals, but I suspect this one may be out of my league.


