It might not be logical to ask others to find a counter-example IMHO during a test/interview... Because we are problem solvers and are supposed to prove our solution, other than asking our clients/interviewers to find an example to negate it...

I think @zmyzz 's question is legitimate because preorder/postorder itself is not sufficient to represent a general binary tree. (But it works for BST.)

Even though using a # for null makes the representation different from a typical preorder/postorder, it doesn't necessarily prove its uniqueness.

There are three ways I'm sure of their uniqueness:

null for null node + level transversal (Leetcode way) Newick representation inorder + preorder/postorder@xiaoyu.bai might also be helpful here.