[step:Prove (iii) by showing $p$ appears with equal multiplicity in numerator and denominator]
Fix a prime $p$ with $\frac{2n}{3} < p \leq n$. The hypothesis yields two bounds:
\begin{align*}
p \leq n \quad\text{and}\quad 2p \leq 2n < 3p,
\end{align*}
so $2n / p \in [2, 3)$, giving $\lfloor 2n / p \rfloor = 2$. Also $n/p \in [1, n/p]$ with $n/p \geq 1$ (since $p \leq n$) and $n/p < 3/2$ (since $p > 2n/3$ gives $n/p < 3/2$), so $\lfloor n/p \rfloor = 1$.
The multiples of $p$ in $\{1, \ldots, 2n\}$ are $p$ and $2p$; the multiples of $p$ in $\{1, \ldots, n\}$ are just $p$.
We now rule out higher powers of $p$. Since $p > 2n/3 \geq 2/3$, and $n \geq 1$, we have $p \geq 1$; more substantively, $p^2 > (2n/3)^2 = 4n^2 / 9$. For $n \geq 2$, this gives $p^2 > 4n^2/9 \geq 8n/9$; and since $p \geq 2$ whenever $p$ is prime, $p^2 \geq 2p > 4n/3 > n$, so $p^2 > n$. More directly: $p \geq 2$ and $p > 2n/3$ give $p^2 \geq 2p > 4n/3 \geq 2n$ provided $n \geq 3$. We need $p^2 > 2n$: from $p > 2n/3$ we have $p^2 > 2n \cdot p / 3 \geq 2n \cdot 2 / 3 = 4n/3$. For $p^2 > 2n$, write $p > 2n/3$ and use $p \geq 2$: then $p^2 \geq 2p > 4n/3$, which exceeds $2n$ iff $n < 0$ — so this estimate is insufficient on its own. We argue instead directly: if $p^2 \leq 2n$, then $p \leq \sqrt{2n}$, so $2n/3 < p \leq \sqrt{2n}$ forces $(2n/3)^2 < 2n$, i.e., $4n^2/9 < 2n$, i.e., $n < 9/2$. Thus for $n \geq 5$, we have $p^2 > 2n$; the remaining cases $n \in \{2, 3, 4\}$ are checked by hand (the range $(2n/3, n]$ is empty of primes for $n = 2$, contains only $p = 3$ for $n = 3$ with $p^2 = 9 > 6 = 2n$, and is empty of primes for $n = 4$ since $(8/3, 4] \cap \mathrm{primes} = \emptyset$). In all cases, $p^2 > 2n$, so $\lfloor 2n/p^2 \rfloor = \lfloor n/p^2 \rfloor = 0$.
Therefore
\begin{align*}
\nu_p((2n)!) &= \sum_{i \geq 1} \left\lfloor \frac{2n}{p^i} \right\rfloor = 2 + 0 + 0 + \cdots = 2, \\
\nu_p(n!) &= \sum_{i \geq 1} \left\lfloor \frac{n}{p^i} \right\rfloor = 1 + 0 + 0 + \cdots = 1,
\end{align*}
where we used [Legendre's formula for prime-power valuation of factorials](/theorems/???). Substituting,
\begin{align*}
\nu_p(N) = \nu_p((2n)!) - 2\nu_p(n!) = 2 - 2 \cdot 1 = 0,
\end{align*}
so $p \nmid N$, i.e., $\gcd(p, N) = 1$, proving (iii).
[/step]