That's one of those weird things for me. I want some stuff pruned, but a lost GCD per rotation or rotational string can knock more than just a widely used rotation out of whack. My 2.33, 2.4, and 2.44 GCD rotations from level 50, for instance, each fit perfectly to a certain modified rotation with Heavy Thrust of course in that mix, and the extra 1-2 Gierskoguls from a no-Phb ST-focused AoE rotation at 2.33 at level 60 does as well. Moreover, attaching Heavy Thrust instead into the Impulse Drive combo just means doubling up, effectively on Disembowel, but now having even fewer options in regards to that damage boost. You can't pop it early in preparation for AoE. You can't skip DoTs in favor of Doom Spike spam. You can't even go directly into a Full Thrust combo on a mob that's only going to last 4 GCDs. Locking down such tremendous bonuses to a single combo that already far outweighs its alternatives in most, but not yet all, situations, would likely kill that job for me. I'd rather see potency increases across the board and Heavy Thrust removed than that, and would rather keep it in some form than either. I feel like any changes in these directions need to take into account a broad spread of actual, situationally optimal or potentially (with faint rebalancing) optiomal rotational possibilities. Even seemingly blank space can make things fit far more satisfyingly. A lot of that can be reworked to fit as well without that space, but not all, and I don't think we should ever jump the gun for a thin concept if it may have negative actual gameplay impact.
The most obvious bloat for Dragoon to me is simply Fang & Claw and Whirling Thrust, identical abilities apart from their positional bonuses. That's stupidly redundant. Moreover, only the one that procs is ever even available; you can simply macro them together and, apart from having to spam them as the GCD finishes because we can't queue macroed skills correctly in this game, never again notice that the ability was actually split in two, apart from its RNG-based animation. It's like they ran out of ideas for what to do with the level 58 ability and said "if we limit this first one to this side, and toss in some RNG, we can double it over as a second ability from the back." Unless each can have actual unique and interesting effects, there's just no purpose for us to be spending two whole abilities on this. Heck, if the RNG component stays, you can give each position an effect bonus for a particular side while none for the other, and the same tiny bit of randomness we have now stays. Or you can give it bonus potency for just the one position, but keep the positional effect bonuses for each, so that you can choose the compromise if necessary, sacrificing 90 potency for the effect you really want. (I'd imagine that'd have a less positive reception, however.)
Just making abilities easily modifiable via right-clicks with bars unlocked or via the Abilities & Traits pane to swap between mouse-over, on-target, and set-and-click, and removing the arbitrary penalties on macros would go a long way to reduce bloat, much like the combo-button system this community's been suggesting on and off pretty much since beta. If the game's not complex enough for a one-fits-all-situations macro not to be less than ideal, that's something they ought to be working on directly, not simply punishing macros for. (And this is coming from someone who doesn't even use them that often, let alone depend on them in any way.)