I don't understand how using group-theory language to describe number-theoretic properties provides extra insight in this case (e.g. conjecture: all perfect numbers are even is more concise than the group-theoretic description given in the page). Can you expand on why you believe the tools of group theory have something to say about this?
(e.g. for polynomial roots, the connection with symmetry groups comes from symmetries of factorized polynomials, while there's no obvious-to-me connection here as there is no unique-up-to-symmetry integer factorization)
I just found it interesting that certain problems in number theory could be rephrased as problems about cyclic groups. Maybe it could potentially make some easier to solve.