Hi Ron!
I told you that it was better to wait after Lucca to talk about your new essay. This place was deserted until yesterday!
These are my observations:
1) About the first part, the description of "Story before", are you using different definitions of Partecipationism, Illusionism and Railroading instead of the ones in the Glossary? In this essay you talk of the difference as if it was based on the player's knowledge of being in a "story before" situation, and not (as in the Glossary) on the use of the "Black veil" by the GM.
This different is significant, because in my opinion, if we use the glossary definitions, the most functional type is not partecipationism, but Illusionism with the players "in the know" about what it's really happening. The black veil used not as a way of betray their trust but as a way, for everybody, to avoid calling out the illusion of reality. Everybody a partner in crime, everybody invested in keeping the secret. But in this essay you seem to have canceled this kind of Illusionism by considering this the same as partecipationism (where the GM tell them right in the nose that they are pawn and they have to accept it openly, the hell with the illusion and the dream)
2) About part II: it's ironic that looking back to my history of play, I find that way of using "setting" (choosing by myself a little part of it, having players characters fully integrated in the setting, tied to the existing social and geographic conflicts) used by me in different situation, and always with rather good effect. I am thinking for example of my very first D&D adventures I GMed, based on the AD&D module "Under Illefarm" by Steve Perrin (where the PC where all native of a little village with problematic neighbors) and the first part of my try of playing in Glorantha (the PCs were all members of a Rhino Riders tribe in Prax, having to bear with tribal affairs like initiation rites, migrations, raiding, etc, and with a ready-play conflict already in motion by having a male PC being the son of the chief of the tribe and a female PC the daughter of the tribe witch-woman, the other political power around)
The ironic part is that, betrayed as always by stupid "advices" in magazines and manuals, and by the moronic habits ingrained in the subculture, I considered every time these good results as "proof" that the group was "ready" to go to the next phase, to "bigger things"... and I always moved them away from their home setting to play "the better modules", with disastrous results. And I did see that not working at all, but I thought that my group was the problem, because "everybody else" enjoyed these "big story modules". Or at least, that was what was written in the magazines... (I would like to read a real survey, someday, about how many people really played Dragonlance. I personally don't know anybody who got to the end of the second module. Of 16. But at the time it seemed as everybody was playing it)
3) About Part III: see my previous comment. I have to say that these moments that I described were for the most part not "story now" games, so I have not a lot to say about using these techniques in real play. For the most part they were story-after and gamists, but they worked, using the general advice written in part II. It was trying to turn them into story-before in a bigger chunk of the setting (as advised by the rpg culture of the time) that ruined them.
4) about part 4: I still have not found the time to read Degenesis (I returned from Lucca with a lot to read)
5) about part 5 and 6:
The Megaplot:
the general widespread result (and this was the single perceived successful way to publish and recommended way to play in gamer culture by the early 1990s) was to transform a hobby of gamers and nascent designers into a mere captive market of periodical buyers and readers.
Probably that was the plan, though: to sell stuff to people even if they had not the time, chance or desire to play it...
About the principal culprit of all this, I already said by email that I consider TSR by far the first and worst offender (the time of the avatars was the first setting-wide megaplot, and it preceded AD&D2, not counting Dragonlance because in that case the setting came from the megaplot and not the other way around) , but that part I will post on the forge as you asked, seeing that the discussion there is going right there, and it's rather tangential to your essay.
To Alessandro:
well basically all traditional RPGs have a rich setting that could be put to good use following that structure.
I loved to see Unknown Armies being mentioned though 
No, a lot of traditional games have "big settings" in the sense that they have a really big list of monsters and npcs to fight against (or for), a patchwork of different micro-settings for when the players get bored (or to attract more players offering a lot of setting in one box) maps, etc.
But the "setting" in the book is not the "setting" in the Big Model. The first one is a physical product, the second one is a component of play. Often the second one is really much smaller, and it doesn't take a lot from the big list in the first one. Or the setting described in the book is bland and without any thematic hook. The examples are the biggest selling "setting" of the rpg market: where can you find anything to create theme in the Forgotten Realms? Or Dragonlance? With all these "good empires vs evil empires" tripe? (it's not surprising that the themes of the Dragonlance Novels are totally personal: personal duty vs personal honor or love, self-respect, revenge, power... it can be debated if they are treated in a mature manner or in a juvenile one, but there is theme in the Raistlin character story arc. But the setting doesn't enter in these themes, ever). Ravenloft was a INTENDED patchwork of different isolated "isles of terror". Other examples are the World of Darkness, Star Wars, etc.
If you know Glorantha, the difference is striking. And I am not talking only about Hero Wars. Even in 1980, with a system like Runequest that was dedicated to "simulation" above all else (and totally unsuitable to story now play), you did get Glorantha supplements where the "Bad guys" of one culture were the heroes of the other.
At the beginning, Lunars vs Orlanthy. Easy enough, it's celts vs romans with some fantasy trappings, right? (I am talking about first sight, I know the Lunars are Persians, Ottomans, the USA and Christians too, sometimes and somewhere, but let' keep this simple, right?). But then they go with... the point of view of Trolls? The Darkness creatures? (after reading "Trollpack" I wanted to play a troll for years. Never got to, the others wanted to play only D&D. Bah...). And then... CHAOTIC CREATURES? BROOS?
Are they out of their fucking mind? You go in the middle of the worst cesspool of chaos and evil in the entire world, Dorastor, and it's ruled by a... noble and refined broo?
As you can tell, with Glorantha it was love at first sight. And "sight" is the right word. I was never able to play in it. I only got to have a campaign based on it by being the GM, and with nobody other reading a single page of the books, it was hopeless. Sigh(t)!
Anyway, before I digress too much: the difference is really visible, evident, between a "setting" (list of meaningless bits) and a Setting, like Glorantha and Near (first version, the Shadow of Yesterday one). And in this industry the second type is very, very rare.
