I was wondering if there was another way to derive the inequality shown in the title of this question apart from calculus.
Doing it with calculus is pretty straightforward:
Let $f:(0,1)\to\mathbb R$ be a function defined by $f(p)=p(1-p)$.
Taking the first derivative $f'(p) = 1-2p=0\Rightarrow p=1/2$ is a critical point. Then the maximum is $1/4$ and it is reached when $p=1/2$. Hence $p(1-p)\leq 1/4,$ $\forall p\in(0,1)$.
Substitute $p=\frac{1}{2}+q$. The inequality then becomes
$$\left( \frac{1}{2}+q \right) \left( \frac{1}{2}-q \right)=\frac{1}{4}-q^2 \leq \frac{1}{4}, $$ which is obviously true.