Klein's proposal can also be used on arguments: suppose your argument will be found unconvincing (alternatively, suppose it's actually wrong). What were its weak points? Unfortunately my weak point forecasting is ... weak, so when back-and-forth is an option I much prefer that.