I have to ask - and I can't quite believe I'm about to type this - how can you possibly balance those in the same game? I don't think you can, thus either one has to accept imbalance in the game or one of those classes has to go.
You can't balance it, and shouldn't even try. Balance only matters for things intended to be used together. A "mythic" fighter and a "mundane" fighter aren't intended to be used together. (Or more precisely, if you decide to use them together because it fits some niche thing you are doing, or because maybe one guy wants the challenge of a mundane class in a mythic setting--then the balance issues are all on you. That's no different than if you decide to mix, say, 1st level PCs with 10th level PCs. It's outside the tolerance levels of the design.)
And that brings up the next major objection that has to be addressed, and why it needs to be this way: How do you communicate to the player community at large, and then within groups, what things work together? Label those suckers!
For example, you should be able to say something like, "If it says 'wahoo' or 'mythic' on it, you can't play it in my game unless you get express permission--and don't hold your breathe waiting."
This is where the diversity moves from a weakness or neutral thing to a real strength. Because having to satisfy multiple playstyles is difficult. But trying to shoehorn multiple playstyles into a single element is really hard. And to be fair, I think this is part of the visceral reaction that Next is producing--something like: "Hey, they are gonna slip some of that wahoo stuff in the fighter class to try to please those other guys, which is probably going to just botch the class for me--and not even be enough to make them happy anyway."
And that's absolutely true if they try to mush it altogether to make it roughly balanced. One guy likes cabbage and another guy likes pecan vanilla ice cream. Just put cabbage in with the pecans in the ice cream, and it will make everyone happy?
Oh no, some people only want one or the other. OK, I know. We'll put both in a big box in unmarked containers so that you have to open each one and try it to see what it is! It will need to go into the freezer section, but what could go wrong?
This point goes double for feats, spells, etc. Classes will presumably be a short enough list that you can determine pretty fast which ones you want to pay attention to and which ones you don't. The main benefit to the labels in the classes is that if they bother to label, say, the fighter "mundane" and the warrior "mythic," then you can pretty much bet that the fighter really is mundane.
Yes, but I suspect a lot of the poorer add-ons to earlier editions were just that: add-ons, dreamed up long after the original core was released and thus less likely to work with the design of said core. This in fact goes all the way back to 1e - the original UA was a collection of various post-design add-ons; some were good, some were awful.
Here with 5e they have a chance - realistically, for the first time ever - to design the whole lot all at once and simply release it in stages. And the first release should be the core nuts-and-bolts framework with everything stripped to the bone. For some, that'll be all the game they need. Others might wait for some add-ons (I suspect once 5e releases we'll be up to our ears in supplements within the first year, it won't be a long wait) and go from there. To us as consumers they'll all look like add-ons, but in fact they're all part of the core design.
Like I said earlier, I'm fine with a variety of packages and presentation techniques, as long as key modules are designed
and tested early--before the core becomes set in stone.
