Variations in the successor function from Peano's axioms
It works ...
We have to note that according to Peano's axioms the symbol $1$ is introduced as the name for the successor of $0$; thus, according to your approach, we have to use $1$ to denote the "real" number $2$ (the "new" successor of $0$).
Consider that Peano's axioms include the recursive axioms for sum and product.
For sum :
$\forall x(x+0=x)$
$\forall x \forall y (x+S(y)=S(x+y))$.
Thus, applying the second axiom with e.g. $x :=1$ and $y := 0$, we have that :
$1+S(0)=S(1)=4$,
because in the "reformed" sequence : $0,2,1,4,3,\ldots$, the successor of $1$ is $4$.
Consider now :
$2+S(0)=S(0)+S(0)$,
because $2$ is the "new" successor of $0$; thus :
$2+S(0)=S(0)+S(0)=S(S(0)+0)=S(S(0))=S(2)=1$,
because $1$ is the successor of $2$.
Thus, your proposed "reform" seems consistent, because the operation of "adding to $n$ the successor of $0$" still produces the successor of $n$.
Now, there is an easy way to avoid this apparent confusion : we can introduce new symbols :
$1^*$ for the successor of $0$, i.e. as a new "name" for $2$,
$2^*$ for the successor of the successor of $0$, i.e. for the successor of $2$, i.e. as a new "name" for $1$,
$3^*$ for the successor of the successor of the successor of $0$, i.e. for the successor of $1$, i.e. as a new "name" for $4$,
and so on.
In this way, we have relabeleld the "reformed" sequence : $0,2,1,4,3,\ldots$ with a sequence of new "names" : $0, 1^*, 2^*,3^*,\ldots$.
Using them in the above "computations", we get :
$2^*+S(0)=2^* + 1^*=3^*$
and it is correct, because $3^*$ is the new "name" for $4$.
In the same way :
$1^*+S(0)=1^*+1^*=2^*$
and again it is correct, because $2^*$ is the new "name" for $1$.
Conclusion : the proposed "reform" us useless.
According to Peano's axioms the "real" number $1$ has only one "relevant" property : to be the (unique) successor of $0$, i.e. the successor of the unique number without successor.
As long as we satisfy the two basic property of the successor function :
$0$ has no successor
no two different numbers can have the same successor
there are no "metaphysical" properties that can distinguish two numbers for each other if not their "relative position" with respect to $0$.