for pve trash tank... most certainly.... but for pvp I'd rather have 3 points in infected....
And for PvE bosses it doesn't matter....

Is that taking Windfury haste into account?

Yes. But you can put only 2/3 and the stack will lose with a very very low chance.

Supposing you don't wear a lot of hit gear/ expertise gear, so you will have a 10% total chance to miss the attack.

In 12 seconds you will do 2 mangle and about 5 maul. So 7 special. (((1-0.9*0.66))^7)= 0.2% chance to miss the stack.

With only 1 point in infected wound: (((1-0.9*0.33))^7)= 8.5% chance to miss the stack.

I found 2 error in the code I posted at page 24 (now I've corrected it):
- FB scaling was wrong
- KoJ didn't add energy.

Corrected and redo the simulation, dps increase, the best cycle stay the same.

DPS: 3518.91, White: 997.882, Yellow: 2521.03, Bleed debuff uptime: 59.5387%, Mangle debuff uptime: 78.7746%, Powershifts each 5.04202 seconds (average)

Using 5RIP will make a slight increase in dps but you'll need to powershift 25% more, it's basically not sostenible as a cycle.

I was having a play around with the talent calculators earlier, and putting fewer points in IW is basically the solution I came up with as well. For end game raid tanking, I'll be looking at something like:
Talent Calculator - World of Warcraft

The salient points being:

o 2/3 Survival of the Fittest, with the assumption that boss vs player crit chance doesn't change (remains at 5.6%) and that we can pick up the incidental 1.6% required from shared drops (neck, rings, cape etc) and enchants. If so, the only loss is 2% in base attributes.

o 2/3 Infected Wounds which should result in a reasonably high uptime of the debuff, as calculated above.

o No Imp LotP T.T

Alternately:
Talent Calculator - World of Warcraft

If 3/3 SotF ends up being vital, we could easily drop 1 from Feral Instinct and only take a minor hit on AoE threat.

However, IW is 0 effective DPS. And I'm fairly certain (caveat emptor) that having a warrior thunderclap every 30 seconds is going to be a lower DPS reduction and better Mit for the tank than 2 IW

Did a bunch more testing of the energy regen... Very interesting stuff, will post my findings tomorrow. Short version: Wolfshead is working, it just isn't displaying the energy from it, client-side. Server side, it is working. Powershift, you appear to have 40 energy, after 1sec GCD you appear to have 50energy. But Mangle, it drops you to 41. You'd expect it to drop you to 21 without Wolfshead (and it does, without wolfshead).

This will more than likely be the dps build http://www.mmo-champion.com/talent/?...00000000000000 and the tank build http://www.mmo-champion.com/talent/?...00000000000000 depending on what your role is in the raid you would of course switch some things up.

It would be nice if Natural Shapeshifter/Master Shapeshifter was thinned from 5 pts total to possibly 3. Those 2 extra points could be used to fill out IW.

Actually:
Talent Calculator - World of Warcraft

It would be better if Furor was made a base talent, and Natural Shapeshifting Moved from Tier 2 to Tier 1
Tier 2 should Improved Shifting 2pts. 1 point adding 5 energy / 2 rage per shift / Improving Mana Regeneration while in Aqua, Travel, Cat, Dire Bear Form by 10/20%. Natural Shapeshifter should be 2 pts talent reducing HT by 0.25/0.25 and increasing melee damage by 5%/5%

Don't forget Curse of Weakness and Icy Touch

CoW will cost the raid either CoE, Cor or hurts the locks on DPS w/ CoD/CoA. Highly dependent on how many locks are present and what spec they are.

Icy Touch on the other hand may or may not be in a DK's attack rotation anyways, haven't been paying attention to their DPS rotations.

If you had 0 latency, you'd just always powershift at 10 energy. However, there's latency between when you get enough energy to use whatever skill, and when the server actually gets and responds to that request to use that skill, then latency again for when the GCD ends and the server responds to your request to powershift. For example, powershifting after a 29 energy Mangle, with a 200ms ping would look like this:

0.0sec: Energy regens to 29. You send the request to cast Mangle.
0.2sec: Energy regens to 31. You receive the cast confirmation from the server, dropping energy to 2.
1.2sec: GCD ends, energy regens to 12. You send the request to Powershift.
1.4sec: Energy regens to 14. You receive the powershift confirmation from the server, dropping energy to 0.
1.9sec: Energy regens to 5. You receive the Furor notification from the server, bumping your energy to 45. (Note that when Furor applies still varies, but it's always before the powershift GCD ends, so it doesn't really matter)

So net gain from the Powershift is 26 energy for 200ms ping. Basically, the fluid energy change removes how much energy you have after a cast from the equation, and simplifies the energy gain of a powershift to (30 - 2xPING). Powershifting macros should just be set to powershift when energy is less than or equal to (11+PING). 10+PING would be ideal, but pings fluctuate a bit, so use 11 to add a 1 energy buffer.