The problem with the hamburger method is that it is recognizably dishonest. Everyone knows the structure by now, everyone knows that the positives at either end are just there to cushion the negative. So positive feedback in a hamburger will not only be undervalued, it won't be believed at all.