@Peter I disagree with the claim that this question should be closed and deleted (disclaimer: I've posted an answer to that question). Contra an initial comment, it is based on a precise technical result which is very easy to misunderstand; I think that it would be difficult to phrase the question better without already knowing how to answer it (a common problem for foundations-related questions), and I'm curious what concrete improvements are being looked for.
Let $q$ be an (odd) prime, and let $\gcd(q,n)=1$.
Denote the classical sum of divisors of the positive integer $x$ by $\sigma(x)=\sigma_1(x)$.
A number $N$ is said to be perfect if $\sigma(N)=2N$.
Here is my question:
If $\gcd(n^2, \sigma(n^2)) = q\sigma(n^2) - 2(q - 1) n^2$, does it follow that...