TO, mostly, does not feel right at all, because in this example, as usual, its optimised for one single axis of load, while failure of a bike stem can arrive from any direction during an event which is impossible to predict. Therefore, a simple round tube of maximum reasonable diameter would be the most optimal solution
ngl topology optimization isnt just cool but it just feels right