Breaking the Mould

There is no reason that 4e character classes could not be designed radically differently. WOTC has a clear template for them (a flawed one) and there’s a knee-jerk instinct to follow that same pattern, but I don’t think anything makes that necessary.

I’d suggest that all that is really required for a functional 4e class is that it have about the right number of hit points, about the right defenses, and about the right range of attacks and abilities. It’s not an exact science, but it’s not hard to ballpark. If all you need to do is hit those benchmarks, then it becomes much less important how you do so.
For example, you could greatly simplify the game by using each existing class as a means of creating numerous more streamlined sub-classes, by simply taking away choices. The Iron Tempest class begins with these two at wills, this encounter power and this daily, and the power choices (and other choices, perhaps) might be mapped out from level 1 to level 30.
On the other hand, this means that much more radical (or perhaps regressive) ideas could be supported. There’s nothing that keeps a 4e character from gaining abilities as he levels, the way that characters did in 3e. It need not even be encounter or daily powers. While those are the norm, it’s still entirely possible to adjudicate something that’s usable 3 times per day or the like.
Yes, any such model hits the big wall of all third party class creation – it doesn’t work with the character builder. That’s frustrating. But it occurs to me that it’s only a barrier because of the complexity of handling 4e characters as is. If you’re going to change the way you think about classes, is there any reason your new vision needs to be so complicated that it requires a tool?
It’s an open door. If it was only a little bit open before, I think it’s safe to say that Gamma World kicked it wide. No reason not to see where it goes.