It feels to me like being shown a similar shape (and specifically being set a shape-based task) would explicitly reinforce the perspective-interpreted shape, ju

It feels to me like being shown a similar shape (and specifically being set a shape-based task) would explicitly reinforce the perspective-interpreted shape, just like someone presenting you with the classic duck/rabbit image and saying the word ‘duck’ in your ear. Therefore I’m not completely sure this proves you would naturally hold on to that representation by default out in the real world.