Some general thoughts in response to this thread, which is lighter on hyperbole (and hashtags) than a number of other discussions I've seen today - that's appreciated.
Fury and Heroic Strike
Fury warriors currently have four buttons that are some flavor of "Deal X% weapon damage to a target." One of them generates Rage. Two of them have a base cost of 30 Rage. Two of them are only usable sometimes. One of them is off the GCD. But those are nuances, and at a basic level there's a significant amount of overlap. (And this isn't even counting Colossus Smash.) That both qualifies as what we'd generally consider "bloat" and it also makes the spec much less intuitive to figure out without doing some math (or, more likely, consulting a guide compiled by someone else who did the math). Fury doesn't need three different single-target rotational Rage spenders in order to be effective and engaging as a spec - for the majority of WoW's history, Fury hasn't had all three of those buttons. We think it's one too many.
Our vision of a successful Fury design certainly is not one where the warrior sits at the Rage cap for protracted periods of time. We're not saying that the current (as of June 30th) Beta iteration of Fury is perfect or final. It's not. In the next build, we have a change to Wild Strike (reduced GCD) that should allow it to more distinctly serve as an outlet for excess Rage, keeping some of the frenetic pace that defines Fury and distinguishes it from the more deliberate and tactical Arms spec.
Without Heroic Strike, Bloodthirst would clearly be your Rage builder, Raging Blow your most efficient spender (when usable), and Wild Strike would be your filler. There's little lost depth there (instead of hitting HS when Rage-capped or during Colossus, as you currently do, you'd use non-Bloodsurge Wild Strikes), and one fewer keybind and ability of which to make sense.
But maybe we're missing something - we're certainly fallible. We'd like you to check out the changes in the next build, give feedback, update your petition websites where applicable, and then we can discuss further. None of this is final.
The Beta Design/Iteration Process
To repeat, none of this is final. If you haven't followed one of our Alpha (or now Beta) cycles before, there's a ton of iteration that happens in a public format, much of which never sees the light of day on live retail servers. Our betas are not a sneak preview or an advertisement - they are a genuine invitation to join us as we wrap up the development and refinement of our game, to give feedback, to help us find bugs, and yes, to witness some thorns and missteps along the way.
This sort of experimentation on our part happens all the time internally, long before we open our doors to public testing. Back in February or March some skills were cut or redesigned entirely, then returned days later after playtesting internally and realizing that we'd made things worse, and not better. Other times we tried experiments (e.g. moving combo points off the target and onto the player) and ended up keeping them (for now). The nature of our class design process is that we work on live data, right alongside quest content, encounter content, and everything else that gets added into each build. We want those changes to be part of the same data branch in order to test them properly in a realistic and relevant setting. But, of course, once we've moved into Alpha and Beta, this means that you get to witness some of those experiments too.
We could make separate internal builds to keep any class changes that aren't 100% final in our eyes from seeing the light of day. That would limit our ability to test them somewhat; it would certainly limit your ability to play with them, and to give us feedback. It would reduce the rate at which we could iterate, and the number of ideas we could try out. It might save some frayed nerves in the short run, but I'm not sure it would make for a better game in the long run.
Some of our experimental changes make it into patch notes, because our Beta patch notes are ultimately a living record of our moment-to-moment changes to the game. If we don't document those changes on our end, they'll be seen via datamining anyway, potentially out of context and giving rise to greater alarm. But just because something shows up in "official" Beta patch notes does not mean it's going to be in the official patch notes when Warlords actually goes live. And those are the ones that count. The goal of some of our changes to a number of classes or specializations is to make them more accessible. Note that this doesn't necessarily mean "simple." We want our classes to have nuance and depth in practice, and inquisitive and competitive players will always ferret out and theorycraft ways to maximize performance. We're not looking to get rid of that behavior, but we are looking to refocus its impact.
A player should not need a guide to figure out a basic rotation for their spec (e.g. "if I have enough Rage to do either, should I prefer to Heroic Strike or Wild Strike?"), or to understand the basic purpose of their abilities, or their value (e.g. keeping Steady Focus from dropping off is pretty much the most important thing for a Marks hunter, but reading its tooltip doesn't tell you that). But the difference between the master and the above-average player lies in none of those things: it lies in more subtle timing, cooldown usage, interactions of talents, set bonuses, and trinkets, and in an understanding of how to adjust those considerations to suit differing contexts.
Context is essential. There is much focus on training dummy rotations, or theoretical "Patchwerk" rotations for raiders. But in the situations where performance matters, and in which players can truly exhibit mastery, there is far more multitasking involved. Being the best raiding warrior isn't just about being able to mechanically execute a perfect priority rotation: it's about maintaining your best facsimile of that rotation while avoiding Hisek's Rapid Fire and Ka'roz's Whirl; it's about shifting priorities to maximize damage on specific targets in specific windows that are most vital to your overall raid's success rather than your personal place on damage meters; it's about handling interrupts, stuns, or other needed raid utility while continuing to DPS "optimally"; and so forth. Being the best PvP warrior is far more about positioning, communication, awareness, predicting enemy actions, and quick reaction time, than it is about executing any sort of DPS rotation.
I'd venture so far as to say that when you move away from the training dummies in town and into realistic situations, no one has reached the absolute skill cap in either sphere of gameplay in WoW. A precious few players might come close at times, and may string together a few near-flawless pulls or arena matches, but in the big picture, there's always room for improvement, and removing a single rotational ability or a proc-driving passive is not likely to change that.