You can actually estimate a fight's progression fairly quickly, even during progress. You may not be able to pinpoint it to withing 10 seconds, but with a 3 minute cooldown, you also don't need to be that accurate. Maybe it's just me, but even when we chain-wiped on a progression boss, I sure knew it would only take X more minutes until it's dead. Accommodating for such lengths can be done.

That being said, you are of course correct that it can be difficult in certain situations. If, say, a boss fight divides into "cooldown segments" very tightly, small variations can make or break an extra use. But then again, even using it on cooldown can't always save you from that.

In fact, I think "just use it on CD to be safe" is more of a blanket statement than "think about the flow of the fight beforehand". There are no easy, one-statement solutions for these things, and I'm certainly not trying to provide one. I'm merely illustrating the thoughts behind the decision-making process, but in the end it is up to each individual to make their own.

1)- I don't understand how Subterfuge is stronger than Shadow Focus.
My assumed used of Subterfuge is something like:
You pool energy until nearly max, then vanish. You then use an opener (either garrote or ambush). You then wait until the buff is right about to fall off three seconds after, then use another opener.

Is that correct? The gain here is up to 3 seconds of find weakness. Note that ambush used here would cost 60 energy.

Shadow Focus is:
Your ambush is free, saving you 60 energy

Is subterfuge just simming higher? This is very non-intuitive for me, normally three seconds of find weakness wouldn't be worth all that energy.

2)- By the EP values in the post, you still want to gem primarily agility. A yellow socket, for instance, could be 120 Agi for 440 EP, or it could be an orange socket for 378 EP, because the 120 haste is multiplied by the much smaller haste number. Putting a yellow into that would make it 315, even less. So the comparison would have to incorporate socket bonuses, as it always has. I agree that the secondary stat gems are immensely stronger than they USED to be, but as you wrote it it implies that they are stronger than the primary ones- not generally the case.

We have Subterfuge beating Shadow Focus by a small amount. Like, a tiny amount in a comparison of small DPS gains. If you find you're unable to achieve 2.5s of extra FW and MoS uptime per Vanish, then Shadow Focus would most assuradely be better for you. That said, maybe the assumptions built up around it are wrong / poor, but this one might not be as accurate as it could be until we can model openers (but if I were to guess, I'd say that would sway things towards Subterfuge).

What is your opening sequence for Subterfuge? What about with ShF? What have you tried? Are you assuming 0 CP before the fight?

Re: Subterfuge my guess would be that the optimal opener is going to be either:

(1) pre-med Garotte/Ambush/Ambush since Garotte will trigger SV and Anticipation will mean you don't lose the significant number of additional CPs this opener would generate + you extend FW as long as possible. The upside of this opener is high initial burst. The downside is you end up with no energy: or
(2) pre-med Garotte/SnD/Ambush since you'll get SV (from the bleed) and SnD rolling as well as making your third action one that extends FW; or
(3) pre-med Ambush/Evis/Ambush which probably is the very best burst, has a good chance of an energy return following Evis from RS + prolongs FW and leaves you with a decent number of CPs to burn on SnD or Rupture. All it does not do is get a bleed on nor get SnD rolling straight off.

I think that if you can pre-SnD (and you commonly can) then option 3 is likely to be best with the points from the final ambush then going towards a rupture. Option 3 is also a good choice following any vanish mid fight. If SnD is not on before you open I suspect option 2 might be better.

I was actually asking about the simulations. The ideal opener is probably Prem/Garrote+Hemo/SnD (switch these if you are at 5+ CP after Garrote)+Ambush. If you don't have the energy to get 4 abilities in, then drop Hemo. Also even with FW only affecting Eviscerate Rupture still does higher damage, so you will only ever Evisc once Rupt/SnD are up.

Has anyone actually tested 463 vs Legendarie's yet outside of Simulations?
I have terrible Luck with Drops so far and can't do it myself.
On Wednesday Mogushan Vaults i still wore my Fangs and did okay-ish DPS, but i feel it should be higher for my Gear.
AFAIK Shadowcraft still doesn't count in Proc Decay, so im still on the Fence if 463 Daggers are an upgrade.

DoTs do not update per-tick with personal buffs. This means that Rupture's damage is based off a snapshot of your current stats/buffs. You could also roll that buffed Rupture for an entire fight back when Eviscerate refreshed Rupture.

However, DoTs do update per-tick with target debuffs.

This doesn't justify using Garrote instead of Ambush, since both benefit equally from the MoS buff. Or does Garrote do more damage now? I'll assume it does not.

I'd imagine Ambush->Ambush->Rupture->Ambush->Hemorrhage during MoS->5pt Slice and Dice->Shadow Dance once FW wears off will be ideal.

NB, this is assuming DoTs and buffs/debuffs work the same way they did in Cataclysm. I haven't been keeping up to date with MoP changes.

Could the [Ghost Iron Dragonling] be put into the trinket section as well for comparison?

I ran into two problems when I tried to get the numbers on that luckydo:

1) I didn't have the trinket, and I hadn't seen any dummy-smack logs from anyone who had, so I don't know the exact (or even approximate) behavior of the proc.

2) I don't know how to quantify the EP value of DPS.

If anyone can help with either/both of those obstacles, I'd be happy to spit out some numbers for it.

I've heard that Shuriken Toss as a CP-builder works with subtlety as well, not just mutilate. Has anyone tried this?