On consistency, I think I'm a bit of an outlier. But am happy to discuss more and work out if that's really the case.
Some of what is being said about consistency - say, damage dice for weapons and spells; or similar DCs for similar tasks - seems to me to not really get above the level of playing the game by the rules. So I don't see that so much as a virtue for a GM but more like the minimum required to actually play a game together.
Perhaps there are a lot of "games" out there that don't, or that barely, reach that threshold? In which case maybe I'm being too cavalier in regard to it.
But anyway, some of the other things being said about consistency - say, how draconians will work - seem to me to apply mostly to puzzle-solving play. (Eg the players are expected to work out that this particular monster type poses this particular challenge, and then to use that knowledge to help defeat the monster.)
But if we move to another play paradigm, then there's no particular reason why things have to work out the same every time. Eg maybe this time, when the monster dies, the GM spends a resource (eg a die from the Doom Pool) to generate and impose a particular consequence, like that the monster explodes. Of course it should follow from the fiction in some sense, if the game is to be fun and coherent as a shares fiction; but that seems a weaker and more pliable constraint that what some are saying about consistency.
Perhaps I'm missing something, or misunderstanding?