In all situations except dying to burst damage - which really shouldn't be happening - +% healing is basically just as good as -% damage. So taking a +7.5% per point healing talent instead of a 3% per point damage reduction talent seems like a pretty good idea in the large. And the runspeed is useful for any fight where you have to move, which would be, basically, all of them, so Quickening gives you a DPS increase as well as better survivability. Seems to me that makes it better in more or less all possible circumstances.

536 haste seems quite low and 2400 mastery seems quite high (likely sacrificing agi or not using high EP valued pieces). It wasn't until 5% haste unbuffed where I was starting to feel limber in my rotation. About starved on Energy though, Mut doesn't spam GCDs like Combat does. Having extra GCDs is normal and often gives you breathing room for defensive cooldowns or rotation min maxing. If you feel rushed to refresh rupture/SnD then the culprit is likely poor energy usage. Pushing for extra envenoms might serve your rotation better simply by pooling and then belting out specials under the envenom buff.

I tried going heavy haste a few weeks back simply because in sub-bis gear levels haste had a higher EP value for me, but ultimately the values don't change outside the priority Aldriana has listed, and tightening up my play while going back to our normal stat priorities netted me far more dps gains and top WoL rankings than I ever thought I would gain by swapping around to experimental stat setups.

Note that the value of pre-dodge cap expertise is doubled during any time you spend in front of the boss, as expertise also mitigates parry at equal rate to mitigating dodge. If your guild's strategy for a boss involves putting you in front for a sizeable amount of time, for instance, than for that fight expertise should outperform all other secondary stats up to dodge cap.

In truth, I am not convinced of the veracity of this sentiment. The best models in the world compare average results, because that is what results in the best damage. But when you are actually there in the fight, with 5 combo points and energy pooling and 2 seconds left on slice and dice and 1.5 seconds left on envenom, you are now in a situation where what used to definitely be your optimal strategy (refresh after the envenom buff falls) carries with it a large penalty with a low probability (dodged envenom from uncapped expertise and the slice and dice falling). In this case, you "buy" a simpler rotation with a smaller amount of variability by adding expertise- but your damage will fall in the general case.

I do generally agree with your sentiment towards "my haste doesn't feel right", because even with no haste, mutilate should be able to keep a rotation going. I would simply argue that a spec with less complexity (in his case, more free energy, in my example, less worry about a dodged envenom requiring maintenance) has that as something to be argued for, even if the average case dps is less- gearing suboptimally by model can be defended (for you) if it buys you something you can actually really make use of, and results in you doing better in a raid.

On the other hand, if what's really happening is a reluctance to learn a different rotation or technique, then you have limited yourself as a rogue, and limited your raid's dps. I just don't think it's right to dismiss the other side of the argument, because it's not just math, it's also philosophy. If there was a boss that was really hard that had two buffs that spawned at random times, and getting the buff either halved your damage or doubled it (50% shot of either), then you would have guilds that couldn't get the boss without those buffs trying until they rolled double damage on both of them. Blizzard doesn't make stuff that "lumpy", but the idea is still there.

Scanning through logs near the top on fights like Chimaeron and Argoloth (as close to patchwerk as we have available) one of the things I note is they all have a rupture uptime of ~90% and an envenom buff uptime of ~60%. Just some food for thought as you consider the flow of your rotation. Obviously having significantly less rupture uptime will cripple your energy income and cause you to do less of everything.