I think you have perfectly adequately demonstrated that there are specific games which both meet
@Manbearcat's criteria AND operate by traditional means. What is unspoken here is what went on at the table. At step 1 how was the objective of the game established, and was there a possibility for it to evolve in different directions depending on what was found in play? Are there no other things that the players could establish or discover that would violate the 2nd principle (everything builds off what came before). Is there any third possibility for the end state of the game, or is it binary?
In a traditional game the answers to my questions would be, respectively "No, what is to be found in play is fixed, this was established when the GM was handed the module 'Ravenloft'." and "No, nothing else can be found in play, the module defines all the relevant things. Anything else that is introduced is either irrelevant, or you are not playing a traditional module anymore.", and finally "No, the possible endpoints are pre-determined by the structure of the module. There may be some variation (who survives for instance) but the eventual outcomes are already a fixed set of possibilities."
Clearly none of these three answers would meet the principles of most narrative "indie" type games. So, we must conclude that, while a description of a specific fictional runthrough of Ravenloft can produce narrative that could come from either type of game, the actual PROCESS OF PLAY of a narrative game would be entirely different from a traditional game. This is the central tenet of my own thesis, you can achieve the same end result, theoretically, but you can't achieve the same game experience. I think most of us play for the experience, not some resulting fiction.