I like the idea overall, but I think you're missing two important issues:
1) What if a group of experienced hardcore gamers want to run a game from 1st to 20th/30th level? Part of what they like is beginning as adventurers in a lethal gritty world, but as hardcore gamers they also want complexity right out of the gate.
2) What about groups with varying levels of player "skill" with the rules? Players who enjoy the simplicity of your proposed Adventurer Tier, who want to keep gaming with the same group after 5th level but are turned off by the increasingly complex systems.
If the next edition of D&D embraces modular design you'd think these two issues would be foremost in the designers' minds.
One of the things I think you keyed into is how BD&D tried to ease players into the rules. I don't think it was wholly successful, but compare to 3e / 4e where there is a huge learning curve to become competent with the rules. D&D needs a basic set, but I think tying rules complexity strictly to experience level is unnecessarily limiting. It's definitely a good idea, but the idea behind modularity is, in part, that you can choose how "deep" rules-wise your game is independent of the fictional narrative.Good points. For the first one, would it be sufficient to make the increased complexity/power part of gaining options? That is, I mean one of the reasons that Basic 1st level characters are relatively gritty is that the wizard has one spell, and the cleric doesn't even have any yet
Totally, the underlying rules should be constant, but there's the question of how much is transparent and required for players to know. Ideally what players need to know would be contained within their class and a basic rules chapter.For the second, I think modular design can be an answer, but only so far. You can make some classes/roles/etc. simplier mechanically, while still having similar power to the more complex ones. But the rules themselves need to be anchored for one campaign, even if the system supports "E6" style and a 4E style with options.
Epic doesn't sell well because its only epic in name and not execution. Its only the same monster types with higher math rather than offering anything palpably new.
I think that some of the example campaign arcs in DMG2, Demonomicon, Underdark etc do a better job of this. And also The Plane Above, with its (admittedly brief) discussion of "journeying into deep myth" (= heroquesting).I think a lot of the problem exists on the DMG/Adventure side of things. I don't think either DMG or DMG2 really does a good job of emphasizing the effect of tier on the landscape of adventure design.
I think this hits the nail on the head. From Heroic to Paragon there's a definite shift in the style of play that was very much built in the system from the ground up. This is less the case for the Paragon to Epic tier, which really does feel like more numbers and higher math. Although I agree with keterys that Epic tier shouldn't be immortal/god-themed; not everyone likes those themes.

(Dungeons & Dragons)
Rulebook featuring "high magic" options, including a host of new spells.