So in theory a haste (sotf haste cap) + mastery build should be the best for output? I am running with one in last weeks and still find it hard to compare with monk's healing or disc's shell healing .

Is there a way to be able to somehow with our current class limitations to stay close to them?

While crit is not consider a stat to go for could perhaps a haste & crit build be better than mastery & spirit or another build?

Do you reckon my choice to go for very low spirit - haste to sotf haste cap and mastery build using 1k spirit flask (if needed depending on number of healers in raid (25man) else using intellect flask) + intellect food be the best choice to be viable in 25man . I had manage to rank in some fights on normal mode e.g. ranked 25 but still as i mentioned i am trying to find a way to stay close to other classes in 25man.

Any feedback or views to the above comments will be good since i want to hear everyone views .

4pc bonus + SotF + 5730 haste. All WG after SM. Minimal Spirit/mastery to haste trading/reforging (well, maybe not minimal, but a lot more acceptable than for the next break point). It's where I found my comfort zone for the last 2 weeks or so.

Even assuming that every WG is SotF boosted, you have a lot less to gain by gearing for 5730. Wild Growth breakpoints without SotF are the same as Efflorescence breakpoints, since they both tick with the same base duration and period. Unfortunately SotF does not boost the Efflorescence HoT, so Efflorescence gains nothing by going to 5730. Also, since we're assuming all WGs are SotF, WG gains relatively less. Instead of going from 8 ticks to 9 ticks, WG goes from 12 ticks to 13 ticks, or an 8.3% gain.

Using the same mastery approximations we've been using, we would say that 2700 mastery is around a 4.5% gain. For an 8.3% gain to WG to be more than a 4.5% gain to every spell, you'd have to be doing 54% of your healing with WG alone. It isn't happening.

Side track conversation:

Doing some math (sparing readers the tedious equations for now), most of our spells will favor Spellpower over Mastery until a specific point. That point is dependent on the spellpower of your weapon but the general formula looks like (40871 = SP,wpn + 1.16865*INT) (1.1025 modifier if you do not use Heart of the Wild). Plug in your spellpower of weapon and solve for INT, this would be your INT unbuffed and w/o the leather 5% buff where 1 point in Mastery Rating matches the gain in 1 point in Spellpower (note spellpower, not INT). [Sample Calculation, 7007 Spellpower (496 ilvl weapon) would require that I have ~29k INT unbuffed before Mastery Rating gain is on par with Spellpower gain]

On the other hand, I said most of our spells because two spells benefit more from Mastery than Spellpower, Wild Growth and the Lifebloom HoT. You would need ~19.5k more Mastery Rating than Spellpower in order for Wild Growth to benefit more from Spellpower and for Lifebloom even more due to the degrees of magnitude.

Regarding Haste
Doing some quick rough approximations with respect to haste breakpoints. If we were to opt for 9k more mastery or 9k more haste to reach the next RJ breakpoint, RJ would benefit more from the mastery ~4%. Whereas Wild Growth would benefit more from 4k Haste to reach the next breakpoint over 4k more Mastery by about 7% more.

Addendum:
The general formula tells you how much more INT you would need than Mastery Rating rather than the total INT alone. It looks like INT is about 5x stronger than Mastery Rating from the start, but since both increase healing as you gain a lot more of one stat (think thousands of a stat) the gain from the other stat gets better. But as Hamlet pointed out there is a sweet spot and the equation above will tell you how far you are from that sweet spot.

I've been nonplussed by SotF from the start. My attempts actually playing with it are limited, admittedly, so I might be curious what very specific observations people might have from practical playing in hard fights. But just pointing out the apparent throughput increase is something easily discernible on paper anyway, and I think isn't very convincing.

First of all, the speedup of WG is a bit deceptive, since you have to cast SM first, wait a GCD, and then cast WG, before the "improved" WG even starts. So even though it is 1.5x stronger, it's in effect even more backloaded than a usual WG. But just on a higher level, I'm surprised people are so sanguine about having their SM tied up for this purpose. I use SM constantly when I need a simple, cheap heal. Putting SM on cooldown tends to be one of the first things I do in nearly any situation if people aren't fully topped off.

SotF comes with a significant "hidden" mana costs.
--Tree Form is 30s of mana-neutral healing.
--When you want to WG, you're now casting [Rejuv, often], SM, WG. Probably much of the time you wouldn't have needed that Rejuv or SM in there otherwise.
--What do you do when you're in between WG casts and the tank is a little low, or one/few raid members want a little heal? Regrowth them? For me, every time damage starts coming in and I need to output something more than basic HoTs, it's SM first before starting the Regrowths. Would consume far more mana if not for that.

So I always wind up in the same conclusion. SotF looked at in a vacuum is a neat bonus. But I doubt it even does enough to justify the mana consumption. And more importantly, Tree Form is too good--I'd need a much more convincing talent to give it up.

I'm sure both of us will spell out our points mathematically when we have time but--wouldn't the equality point between int and mastery depend on your current mastery level also? At a quick glance, I don't see that reflected in your formula. Since the Int and mastery effects multiply together.

Roughly speaking, output goes like
(A+a*Mastery)(B+b*SP)

(A is 1.1, a is the mastery coefficient 1/48000, B and b are spell dependent (base value and coefficient) but tend to be roughly in proportion for each spell.

Taking log derivatives, the relative value of a point of mastery is 1/(A/a + Mastery) and of a point of SP is 1/(B/b + SP).

Equating those, A/a + mastery = B/b + SP.

A/a is 52800 (amount of mastery rating needed to double healing output from what it is at 0 rating)
B/b tends to be around 11000 for most spells, IIRC. (e.g. Rejuvenation - Spell - World of Warcraft)

So I'm inclined to think an SP and a mastery are about equal in value when you have 42000 more SP than mastery. That would make Int better for everyone--but there are few things I've not taken into account that I'll have to come correct for later.

Looking up at your post though, my quick estimate looks very similar to the 40871 you have in yours. So did you mean to say in your equation:
"(40871 = SP,wpn + 1.16865*INT) - Mastery"

If so I think we're both doing roughly the same thing here.

Most of the time people would benefit from WG, there are people that would benefit from SM as well. So SotF or not, you'd be casting both anyway. With the 4pc bonus having the cooldowns aligned, it just feels natural to use SM first, then WG.

Maybe it's just me, but I found it very frustrating to try to fit ToL to different boss mechanics. Yes, it can be mana neutral, but it usually just ends up being a lot of overhealing when used with that intention in mind. And trying to use ToL for it's healing buff instead, would make a lot of sense if you'd also spec for NV and only use them together, but that again would restrict it's usage to certain boss fight patterns.

It seemed to me, while gearing up, that both the increase in spirit and the increase/excess in haste made ToL ['s mana benefit] less apealing then SotF ['s on demand, short-cooldwon, "burst" aoe healing], while not being crazy about either of them.

I agree that ToL is awkward, so I use it as a mana cooldown in place of my normal healing. Instead of the usual rejuv, I'll toss out mass lifeblooms and make up the throughput difference with omen procs for regrowths. I can usually heal through the duration without losing mana, but also not really gaining any either. I haven't put pen and paper to it, but I have a feeling my actual throughput is slightly lower, especially at the beginning of the cooldown versus normal healing for the same duration. Honestly it's not an elegant solution, but it does work.

I've been using SotF a bit more here and there, fights where it's difficult to find any use for ToL, such as heroic Zor'lok (2 healing). The SotF healing style also helps me make better use of my 4 piece and keeps my mastery up time 100%, where without it, I tend to let it drop a little because I'm a miser with my mana (no rejuv's up to SM), or I'm not getting omen procs when I could use them to refresh. SotF is less mana efficient for sure, but in certain situations it's more useful and can also help with keeping things on cooldown.

I use SotF as an attempt to fit the square peg of the resto druid toolkit into the round hole of burst AoE healing. In this I would concur with oopsminded that raidwide AoE damage is something I want to have both Swiftmend and Wild Growth, and ideally even Mushrooms for anyway. The Rejuv requirement isn't a problem either, as you only have to have cast a Rejuv on basically any target in melee or ranged clump in the previous 12s, which I would say not very onerous. I suspect if one ever really needs to burst those two spells to counter frequent raid damage spikes one could try Dream of Cenarius, but I haven't yet encountered anything that would require it.

As for the case of a raid member needing a little heal? If I have a lot of mana I might throw a Rejuv, but more likely I simply do nothing. Barring encounter mechanic issues no one is going to die at 85%, so I don't cast inefficient spells that aren't necessary. I wait, regen mana, and sweep them up with the next out of phase WG. Or also a healing partner takes care of it with their efficient spells. Likewise if my other healer is just better at efficient tank healing, I don't do more than HoTs and Clearcasted RGs until the tank starts taking enough damage to justify larger heals.

I'll concur a bit that ToL always feels awkward, if only because of the conflict between using it frequently for mana savings and using it sparingly to better match the boss at hand.

I'm acctualy using tree of life as a raid cooldown in 10 mans, it's been working really well and I acctualy feel safer using that then Tranq, not entirely sure what the healing numbers are, haven't even looked in to this, but I'd say if you raid 10 man you want Tree of Life, having 10x3 stacks of lifebloom with a stronger wg rolling is always going to be broken. You can't directly copy this to 25 man since you can't, and shouldn't, be rolling lifebloom stacks the same way, but seeing how SotF is even more so a 10 man talent then ToL the choice seems simple to me.

I'd just like to ad in a warning to everyone, there seems to be some weird issues with NV and CC right now, it broke CC on Wind Lord twice out of 42 attempts last night, I'm assuming this was some weird interaction and that it does acctualy avoid breaking CC in normal cases since this is the first time I've seen it happend.

Yes, that's the full version of the calculation I was trying to outline above. Just to tweak a bit:

In practice, people are comparing Int to mastery more often than SP to mastery. And for Int, the SP coefficient is effectively 1.1*1.169 times higher than what the spell tooltip says.

Framing it just slightly differently from how we both did above:

H = (A + a*mastery)*(B + b*(wepSP + c*Int))

A is 1.1625
a is 1/48000
B/b I'm going to call 10000. It's just under 11000 for Rejuv and WG, just over for HT. But then I'm reducing by factor of 1.1 to account for the spellpower buff, so using 10000 instead of 11000.
c is 1.1686 with HotW, 1.1025 without.

Taking logarithmic derivatives (this just means taking the derivative and dividing by the original function, often easier to work with). But essentially the same partial derivatives you used.

dH/dMastery * 1/H = 1/(A/a + mastery) = 1/(55800 + mastery)
dH/dInt * 1/H = c/(B/b + wepSP + c*Int) = c/(10000 + TotalSP)

equating dH/dMastery to dH/dInt:
c*(55800+mastery) = 10000 + TotalSP
With HotW, 1 Int is better than 1 mastery until:
TotalSP = 55211 + 1.169*mastery

Although, what we really care about is when a half Int is better than a mastery. Just drop a factor of 2 in the right place:
c/2*(55800+mastery) = 10000 + TotalSP

TotalSP = 22605 + 0.5843*mastery

That I think is the most useful result. If the left side of above equation is higher, gem mastery, if the right side is higher, gem Int.

-------------

PS, in your post:
"The second interesting observation is that the second term in both equations have the same slope which implies that they would increase the healing done by RJ at the same rate on a 1 SP : 1 Mastery basis. "

Careful, that's not what this means. 1 SP (even without % buffs) increases healing by far more than 1 mastery; this is in fact the very thing we're proving. You've already taken partial derivatives w.r.t. SP and mastery. If you did the arithmetic for the value of those partial derivatives, the SP one would be larger, which is just what you said in the previous paragraph. The fact that the two partial derivatives have the same slope w.r.t. the "other" variable is just because the mixed second-order derivative of your PR function is the same regardless of which order to take the derivatives in (see Symmetry of second derivatives - Wikipedia, the free encyclopedia ).

I'm raiding 10 man though so having tree of life + nature's vigil is pretty much needed for the last phase, especially since we lack cooldowns. Since it happend twice in 42 tries and I'm not even sure that was the cause of the acctual breakage then I'm going to stick to it.