If I am correct, the branching factor is the maximum number of successors of any node
You are correct, they should also be the immediate ones:
If 11 is the goal state and I start going backwards, is 10 considered as successor of 5? Even if it do not leads me further to my start state 1?
No, there is also a bit of misunderstanding of bidirectional search: In bidirectional search you run 2 simultaneous searches, one forward from the initial state, and another one backwards from the goal(hoping they meet in the middle and save you steps), if actions are reversible ( going from node to node), the successor nodes become predecessors in one search and vice versa, and your goal becomes your initial state, in your case:
Artificial Intelligence: A Modern Approach, by S. Russell and P. Norvig.