Just to make trouble and maybe tempt board member Mete or another metallurgist to contribute, I've always heard that sizing corrupts the surface crystal stricture enough to encourage more rapid migration of alloy constituents to a softer arrangement (age softening). Thirty years ago this was commonly given as a reason bullets shot as-cast (where molded and dropped size allowed that), seemed to lead less and be more accurate than sized bullets. The idea is that the induced softening takes longer than barrel time to occur, so it's OK for the bore to size the bullet during firing, and only pre-sizing would cause the problem. Kind of the opposite of work hardening.
The problem with the above is that I don't think I've seen a metallurgist or systematic testing back it up, so it's just more speculation. And there are, of course, other possible explanations. The main one would be that many revolvers shoot better with bullets sized to fit their chamber throats as closely as possible, and lever guns often seem likewise to produce best accuracy with hard cast bullets sized more like 0.002" over groove diameter than the usual 0.001" over groove diameter. And that's closer to what most molds drop. Being oversize will also raise start pressure and engraving pressure a little, which can improve pressure consistency and lower velocity SD. In the 1911 I've always found headspacing on the bullet in the throat leade produced best accuracy and least leading. Getting a cast bullet aligned straight in a bore seems to be more critical to best performance than with jacketed bullets.
The problem with the above is that I don't think I've seen a metallurgist or systematic testing back it up, so it's just more speculation. And there are, of course, other possible explanations. The main one would be that many revolvers shoot better with bullets sized to fit their chamber throats as closely as possible, and lever guns often seem likewise to produce best accuracy with hard cast bullets sized more like 0.002" over groove diameter than the usual 0.001" over groove diameter. And that's closer to what most molds drop. Being oversize will also raise start pressure and engraving pressure a little, which can improve pressure consistency and lower velocity SD. In the 1911 I've always found headspacing on the bullet in the throat leade produced best accuracy and least leading. Getting a cast bullet aligned straight in a bore seems to be more critical to best performance than with jacketed bullets.