When I first read this question (Conjecture about the system of powerful equations) about systems of equations where $(a^n-1)\equiv 0 \mod (b^n-1)$, I assumed the question indicated the modular relationship was to be true for every $n\in \mathbb N$. Reading the comments, however, it became clear that the question was about systems where the modular relationships are true for all $n\le m$.
If $(a^n-1)\equiv 0 \mod (b^n-1)$ for all $n$, $a=b^k$ is one possibility, as that would result in $((b^n)^k-1)\equiv 0 \mod (b^n-1)$ which is true for all $n$. However, for the narrower case where $n\le m$ (in particular $m=2$, so $n=1,2$), an example was given where $(a^n-1)\equiv 0 \mod (b^n-1)$, but $a\ne b^k$ ($a=5,\ b=3$).
This raises two related questions:
1. In the case where $n$ has no upper limit, are there solutions to the system of equations $\forall n: (a^n-1)\equiv 0 \mod (b^n-1)$ other than $a=b^k$?
2. If not, then is there a value of $m$ above which there are no other solutions than $a=b^k$?
Reference for part $1$ see here (the first answer). From my proof (the second answer) we get that it's enough to assume that $m = a$ and it gives some answer for the part $2$. In fact the same proof allows us to formulate the next motivational theorem (this theorem can be seen as a motivation to the Conjecture).