Now, from Eisenstein: "...from preconceived methodological positions, nothing grows. And a tempestuous stream of creative energy,unregulated by method, yields less."*
So, seeing the game system (as opposed to the game played) as the method, we derive that fixing the method rigidly strangles the potential of the art -- the play -- despite the play requiring some method in order to effectively channel creativity into art. And so, the best methods are derived as part of play, using the written method as a springboard. Methods that assume they will be used without modification assume that they will be used to produce a weak art -- something that might be interesting once but must stagnate as the method is "used up". Modification of the method to suit the needs of the art is the only mechanism for progress.
(Adding the analogy: the paint by numbers set provides perfect rigid method for creating precisely one painting. All the tools are present in it to paint a great many paintings by selectively ignoring portions of the method, but a strict reading of the method produces only one painting. Few would recognize the results of a paint by numbers set used strictly, as art.)
Consequently, effective RPG rules are those that admit to change either by design or by accident. Games that are seen as "flawed" compared to either technical manual (rigid method) or fiction (art) nonetheless may be optimally designed to ensure that play is ongoing, changing, deepening art by providing the space to modify method as play gains sophistication or changes direction.
*"A Course in Treatment", _Film Form_, Sergei Eisenstein 1949, Harcourt