If they were discarding old feats as they went, some slow growth wouldn't be bad. Start with 30 feats. Practice shows 5 of them aren't all that. Make up 7 more, deprecate the 5. Next go around, get rid of those 5 completely, evaluate the other 27 again. Add a few more, mark some as borderline. And so on. Eventually, you have a solid core of great feats that you can confidently state, "these we keep." The rest of it is in flux.
If Themes are feat-delivery mechanisms, I don't think that can work. If you deprecate a feat, that means having to errata each theme that used it.
If they're going to errata, I'd rather they errata the bad feat and make it good.
But constant adding of more and more? That's just making it that much harder to find what you need.
If you don't care about varied or optimal character building, then what you "need" is in the core rulebook. Adding more in splatbooks is just benefit to those that actually want those options.
As far as the "super feats," I agree those have problems. I said as much above, when I indicated historical problems with classes, though I guess that might have been oblique. My main point, though, is that whatever problems broad feats have, they are easier to balance and manage than narrow feats. Where to draw the line exactly is the big question.![]()
Not really, unless you permit balancing the broad feats by having tradeoffs within the feat between the pillars. I.e., intentionally making the combat part more potent, at the expense of the other two. But I don't think that's desirable, since it defeats the whole "equal amounts in each pillar" principle.
As long as there's a constraint that each sub-feat be balanced with each other, I don't think there's a meaningful difference in difficulty of balancing the super-feat system, and having those sub-feats split out into feats, skills, and traits.
The only real difference is the greater number of combinations between feats, skills, and traits, and potentially unexpected synergies. But that seems rather unlikely given their separate spheres of operation. I don't think that small gain in design ease is worth the quite large loss of flexibility for players to define their characters as they wish.