Good point, I am sure the MMs do some gamma or reverse gamma scalping to compensate for their possible weekend losses. But then I suppose that would imply that there would be no change in the spread's value (barring a large movement in the underlying and all else held equal) from Thursday to Monday afternoon. If that was consistently the case, then no one (figure of speech) would want to buy/sell these options on Thursday or Friday, they would wait until late Monday or Tuesday morning. This decrease in demand (bringing down IV), would of course decrease the price of the long options and increase the price of the short options, and thus increase the value of the long butterfly. So I think that over the long run, this strategy has to work. The main problem is that will the profit from this trade be enough to compensate us once we include commissions.