The redefinition of feats.

[Crazy Jerome]

I think that this idea (incorporating race and class options into the race/class) isn't a bad one, but a bad one for D&D's publishing model.

D&D doesn't produce all of their races, or classes, at one time. They will introduce new races (which fit into old classes) and new classes (which are available to both old and new races) all the time. Thus, it is untenable to roll the race combinations into the classes (since you don't know all the races when you design the class), or roll the class combinations into the races, etc.

Oh, I agree with you--which is why I said there really isn't a good way to do this. The way they chose was a compromise, though--and I think they would have been better off to pick one side or the other. You'll note that having "elf, beastmaster" feats in the feat list also does not work well when introducing new races or classes that maybe should have had that feat. That is, either the balance is ultra important--so do it the way I said and make things limited. Or the flexibility/combined list is ultra important--so have the combined list, but don't make the feats so restrictive (and there will be balance problems, not matter how much they test). What we got was, "they are both sort of important"--so a solution that is very clunky.

My suggestion of having a separate, orthogonal design space for the multiclassing and feats is recognizing that in the current design--the problem is insolvable. There is no publishing arrangement that will be entirely satsifactory to the competing needs--whether WotC wants to publish a bunch of new races and classes or not. You could drop feats all together (as in making the various abilities simply race or class abilities). You could expand the whole thing into a larger, orthogonal sub system. You could replace it with some other design entirely. Or better yet, do some combination of all three. But leaving a compromise, hodge-podge system in place will result in a compromise, hodge-podge presentation.