Rawr's tree model is very poor. Continue doing what you're doing; don't change your gems around according to what it recommends.

You're computing the average cast time of Nourish? That's not very useful. I think his point was to prevent clipping when NG was active. The way you should handle the buff is simple. Cast time with it active and with it not active; two separate cases.

I agree, Rawr is not 100% correct for resto druids. Stay with gemming for SP.

To have 50% uptime with Nourish chain casting, you'd need to have 1 in 6 casts be crits as each Nourish crit lasts for 3 casts even at 1.25sec cast speed (i.e. no haste rating other than NG). I think it's safe to assume our crit rate is closer to 50%, not 17%, so even if the question was "average nourish cast time", assuming 50% is just wrong. Having 45% crit is no problem when raid buffed (20% from gear/raidbuffs and 25% from talents) and for Nourish to drop off we need 3 casts in a row not to crit, which is a 9.1%~ chance, so assuming a 90% uptime for NG when spamming Nourish feels a lot more real. Assuming you know that the chain casting will be required and you already proc'd NG waiting for the damage to be incoming.

However, back to the matter at hand. 510 haste sounds a lot closer to a haste hard cap number I expected. My next question is what do we do about this? Stacking crit is horrible, we all agree that. Dropping NG to get enough haste (I think it's about 1269 haste using your formula) sounds just as bad, and the gain from the 3 talent points is merely 3 in Tranquil Spirit for me (and a final point plus surplus in imp.tranq/emp.touch from GotEM points). I honestly don't know what gear to pick up any more. My only thought on the matter is that I wished NG always had 100% proc on crits and offered 7/14/20% haste, that way we could drop 1 point and stack more haste slowly, by taking points out of both GotEM and NG as our gear progressed.

With 800+ haste, I can see a build like: (yes, that's 1sec RJ/WG GCD)

Talent Calculator - World of Warcraft

working, but until our haste reaches that level, I can't see anything but just humping our GCDs and having sub-1s nourish casting. It's not like we're not used to sitting on our GCDs anyway. I'm totally out of ideas, I hate to just ask questions like a nub and not offer much in terms of answers but I don't see this answered anywhere on these forums. What is the best move after this haste hard cap?

It does not get any benefit from Imp Rejuvenation, Gift of Nature or Genesis and there is no logical reason for it to get them so working as intended - despite how much many of us would want it to include them into the calculation.

According to maths (I don't know if anyone already gave it), this instant heal from 4T8 seems to be : 0,45*SPELL+340,75

The problem with Rawr's Tree Model seems to be that it weights longevity too heavily vs throughput. If you find the option and adjust it from 50/50 to something like 70/30 throughput, the results become more reasonable.

To not clip the gcd while nourish-spamming, which was the entire point with what he wrote.

[Boots of the Petrified Forest] got buffed in 3.1.2.

It gives you now 93 SP, but little less spirit, 53. Good Haste Option for the Slot.

On the very first page of this thread, there is a breakdown of how Gift of the Earthmother has soft caps as far as haste goes. On the chart "Wrath of Air and Improved Moonkin Aura/Swift Retribution Aura" it lists 10.95% as the soft cap for 5/5 of Gift of the Earthmother.

I am trying to figure out where the missing percent is. Wrath of Air gives 5%, and then Improved Moonkin Aura gives another 3%, for a total off 8%. 10.95% + 5% + 3% = 18.95% ... still 1% shy of the 19,95% needed for a 1 second GCD.

or am I missing something?

ok, thanks... they are multiplicative, not additive, that was my mistake, thank you very much... I assume Celestial Focus is the same way, which suddenly makes the math make sense to me.