$$\begin{align} 2{\sqrt 2}\sin(\phantom{2}x) +3\cos(\phantom{2}y) &= \phantom{-}3.5 \\ 2\sin(2x)+5\cos(2y)&=-0.5 \end{align}$$
I've gotten to the point where I've gotten an equation for $\sin x$ and $\cos x$ using the double angle formulae, but they're horrible roots of quadratics and I don't believe it to be a good method, how else could I attempt this question?
For sure, this is an ugly problem and a numerical method should be used.
What I did was to take $y$ form the first equation $$\cos(y)=\frac{1}{3} \left(\frac{7}{2}-2 \sqrt{2} \sin (x)\right)\implies y=\pm\cos ^{-1}\left(\frac{1}{6} \left(7-4 \sqrt{2} \sin (x)\right)\right)$$ Replace $y$ in the second equation and plot $$f(x)=2\sin(2x)+5\cos(2y)+0.5$$ For the range $0 \leq x \leq 2\pi$, roots look to be close to $x=\frac \pi 4$ and to $x=\pi$. In fact $f\left(\frac{\pi }{4}\right)=0$; so we have one root.
For the second root, we could expand $f(x)$ as a Taylor series built at $x=\pi$ and get $$f(x)=\frac{82}{9}+\left(4+\frac{140 \sqrt{2}}{9}\right) (x-\pi )+\frac{80}{9} (x-\pi )^2+O\left((x-\pi )^3\right)$$ Ignoring the higher order terms and solving the quadratic leads to the approximation $$x = \pi-\frac{5 \sqrt{2}-3}{10} \approx 2.73449$$ while the "exact" solution is $2.71351$.
For sure, we could be better using on more term in the Taylor expansion to get $$f(x)=\frac{82}{9}+\left(4+\frac{140 \sqrt{2}}{9}\right) (x-\pi )+\frac{80}{9} (x-\pi )^2-\frac{2\left(36+35 \sqrt{2}\right)}{27} (x-\pi )^3+O\left((x-\pi )^4\right)$$ and solve this nasty cubic equation (awful expressions for the three real roots) and the solution of interest is $2.70626$ which is better but at the price of a small nightmare.
Using Newton method with $x_0=\pi$ would give the following iterates $$\left( \begin{array}{cc} n & x_n \\ 0 & 3.14159 \\ 1 & 2.79115 \\ 2 & 2.71796 \\ 3 & 2.71353 \\ 4 & 2.71351 \end{array} \right)$$