Flats added to the sides would be flats machined at the spot where one sleeve contacted the next one to allow the bores to be centered/aligned over the crank throws, as the sleeves were too large in diameter to maintain the proper bore spacing. The Darton article I mentioned earlier also ran into this problem : they solved it quite cleverly : after installing the first & third sleeves, then doing the boring for the second & forth the problem was automatically solved by the interference bore into the already installed sleeve to 'clear' the incoming sleeve, & insuring a "snug" fit at the same time. No need to calculate & add an extra machine process if this method is used. THE KISS PRINCIPLE WORKS!!