Taken alone, not bad story material; but in the context of an established setting...blowing it up to that extent is just bad business.
There are very few setting reboots/large-scale changes that have turned out well.
Basically, the pre-4e Forgotten Realms was fairly divisive. While many people liked it, others couldn't get past its perceived flaws such as the patchwork feel, the abundance of high-level NPCs, highly active deities, and constantly advancing timeline changing things all over the place.
So, when doing the 4e version, the designers radically altered the setting, in order to both correct the previously mentioned flaws and to explain the differences in the game (dragonborn, warlocks, eladrin, etc.). The timeline moved up about a century with a magical cataclysm in the meantime to provide a rationale for the changes. The problem was:
1. People who previously disliked the Forgotten Realms were already inclined not to like the new version, particularly since the process of getting to the new version highlighted many of the things they didn't like in the first place (activist deities and radical changes).
2. People who
did like the old FR saw their beloved setting pretty much destroyed and replaced with a new one with some shared geography and names, and invalidating two decades of published material.
I think, if you're going to reboot a setting you should start with one that's failing in the first place, and I'm pretty sure FR wasn't. Rebooting a setting that's partially successful will only lead to no-one liking the reboot. But if you start with a game/setting that's been out of production for a few years, and identify the things that do and don't work with the setting and make sure to keep the first and address the second... then you might succeed.
There's a Swedish game called Mutant (sort of but not exactly related to the current game Mutant: Year Zero) that I think actually provides an example of both approaches. Back in the mid-80s (I think 1984, but I'm not sure) the game Mutant was released as the first professionally published Swedish homegrown RPG (the game Drakar och Demoner preceded it, from the same company, but the first version of DoD was a translation of Basic Roleplaying + Magic World from Worlds of Wonder). The game was an unabashed Gamma World rip-off, but using a BRP-ish system, and having the sketches of a setting called Mutant Scandinavia centered around the Pyri Commonwealth in parts of what is now Sweden. This setting got more detailed though various supplements and support articles written throughout the 80s.
But apparently the company making Mutant wasn't happy with its commercial performance, and they saw the new cool Cyberpunk genre growing and wanted to get in on that. So they scrapped the old Mutant product line, and made a new one, also called Mutant. This was more of a dark future/cyberpunk game (though originally lacking in rules for cyberspace). You would still have mutants around, but they were more like morlocks or the neo-men from Camelot 3000. The focus moved from exploration of Forbidden Zones to dealing with gangs and megacorps in shady mega-cities. This product line was cancelled after only two years, in favor of the more 40K-inspired Mutant RYMD which later evolved into Mutant Chronicles.
Anyway, move up a few more years, to 2002. By now, the previous publishers of Mutant have gone bankrupt and been restructured as Paradox Entertainment (holding the rights to the old games) and Paradox Interactive (making some pretty nice computer games, though they weren't as big in 2002 as they are now). Some dudes in northern Sweden license the rights to Mutant from Paradox Entertainment, and create a a new version of the original Mutant. It still uses a Basic Roleplaying-based engine, but with many improvements on the old. You still have the four primary classes of characters (more like races in a D&D context): Human (non-mutant), Mutant (both mutant animals and humans), Psychic Mutant (also including animals), and Robots. The setting is back to Pyri Commonwealth, but a slightly earlier version, and the description of neighboring areas are more sketchy. The writing and illustrations are top notch at setting the mood, which I would characterize as satirical social-realist - there's a lot of emphasis on the struggles of the oppressed working classes of both mutants and humans, and the human-controlled upper class, and such. It's hard to describe, and very uniquely Swedish. Anyhow, this game became (by Swedish standards) a runaway success, showing that a setting/game reboot
can work. The product line was cancelled after seven years, and the publishers said that they didn't cancel it because it wasn't doing well, but rather because the license was running out and they had already done a core book (and a revision), small splatbooks for the classes (collected into a big players' expansion book), a monster book, a highly acclaimed trilogy of adventures/sourcebooks forming a pretty epic campaign, and assorted other things. Basically, all the low-hanging fruit had already been picked, so it wasn't worth extending the license.
TL;DR: Taking an active, and already beloved setting and revamping everything about it will almost certainly not work. Taking an inactive setting and respectfully reinventing it, keeping the things that work and de-emphasizing the things that don't, might. The Swedish game Mutant provides a good example of both approaches.