Whilst $\text{holomorphic}\circ\text{harmonic}$ might not be harmonic (for instance, $\sin\circ\Re$ is not harmonic), it is true that $\text{harmonic}\circ \text{holomorphic}=\text{harmonic}$.
An algebraic way to prove it: call $\partial f=\frac{\partial f}{\partial z},\ \overline\partial f=\frac{\partial f}{\partial\overline z}$ and $\Delta^c=\partial\overline\partial=\overline\partial\partial=\frac14\Delta=\frac14\frac{\partial^2 }{\partial x^2}+\frac14\frac{\partial^2 }{\partial y^2}$. Recall the identities $\overline\partial\left[ \overline f\right]=\overline{\partial f}$ and the chain rule $$\partial[f\circ g]=\partial f\circ g\cdot \partial g+\overline\partial f\circ g\cdot \partial\overline g\\\overline\partial[f\circ g]=\partial f\circ g\cdot \overline\partial g+\overline\partial f\circ g\cdot \overline\partial\overline g$$
\begin{align}\Delta^c[H\circ f]&=\partial\left[\partial H\circ f\cdot\overline \partial f+\overline \partial H\circ f\cdot\overline\partial\left[\overline f\right]\right]=\\&=\partial\overline\partial f\cdot\partial H\circ f+\overline\partial f\cdot\partial\left[\partial H\circ f\right]+\partial\overline\partial\left[\overline f\right]\cdot\overline\partial H\circ f+\overline\partial\left[\overline f\right]\cdot\partial\left[\overline\partial H\circ f\right]=\\(H\text{ harm})&=\Delta^c f\cdot\partial H\circ f+\overline\partial f\cdot\partial f\cdot\partial^2H\circ f+\overline{\Delta^c f}\cdot\overline\partial H\circ f+\overline{\partial f}\cdot\partial\overline f\cdot\overline\partial^2H\circ f=\\&=\Delta^c f\cdot\partial H\circ f+\overline\partial f\cdot\partial f\cdot\partial^2H\circ f+\overline{\Delta^c f}\cdot\overline\partial H\circ f+\overline{\partial f\cdot\overline\partial f}\cdot\overline\partial^2H\circ f\end{align}
And it's apparent that this quantity is $0$ on any $\Omega$ where $f$ is holomorphic (because the condition of holomorphy is $\overline\partial f(z)=0$ for all $z\in \Omega$).
For "A function which is everywhere locally harmonic is globally harmonic, showing that the statement is true.", why does being locally harmonic imply it is globally harmonic? Is this by Identity Theorem or something? – Mr. Bromwich I Nov 10 '17 at 02:20