On the interpretation of ice-shelf flexure measurements