Is that really so though? If Haste is mana-negative relative to HPS, then Mastery is neutral and Crit is mana-positive. If we broadly consider throughput to be equivalent regardless of secondary weighting, Crit still offers up benefits in the form of reforging spirit into throughput stats.

I think we can do the requisite calculations to see which throughput stat is truly the best and I've tried to below. It would be nice if someone could point out where my calculations/thoughts are wrong. We already have breakpoints for Mastery vs. Haste/Crit, which are admittedly a little bit incomplete when it comes to 4.3. In addition, I'll borrow the methodology from Hamlet/Aanzeijar's post and do some calculations with a pretty standard scenario: AoE, no FI, RT, 12.5% haste, 4 jump CH, using Burning meta. For the purposes of simplification I've made base crit = 0% and base mastery = 0. The spell macro is HR, RT, CH x 4.

+100 Haste is equivalent to 129.22 HPS over the time frame, assuming 6 target Healing Rain. +100 Crit is equivalent to 105.74 HPS. Via Resurgence the MP5 benefit is 31.58. At 7000 Intellect, 31.58 MP5 is worth roughly 45 Spirit. Since we've established that both Haste and Crit translate to > 1 HPS per rating point (even when I pretend to have 0% base crit and 0% crit from Intellect) 45 Spirit, if reforge-able necessarily translates to more than the 23.4 HPS difference.

The calculation is doable for Mastery as well, but we approach it from a different angle. The same MP5 benefit remains. Mastery is more effective than Crit < 65% for AoE, and < 43.5% for single target. If we substitute, point for point, the MP5 benefit converted into Spirit converted into Crit, then those values drop to < 48.8% for Mastery to overtake Crit in AoE situations and < 18% for single targets. Obviously this is best-case and not indicative of anything in the real world. It is important to note, however, that when encounters are considered one by one that the lower values become less and less prevalent, especially after factoring in Maw of the Dragonlord, which will crit and is unaffected by Mastery.

There are a few caveats that change things:
Healing Stream Totem will not crit, but is affected by Mastery. However, the amount of healing that HST accounts for is usually less than Maw in a 25 man setting, and the Mastery portion even lower still (since it scales with health and Crit from Maw does not).
Haste affects neither.
Mastery over time diminishes its own effectiveness. The consequence of this is that it shortens the window, in say, fights like Ultraxion, where Mastery is more effective than Crit at the end.
There is the TC consideration, where Haste > Crit >>> Mastery.

The assumptions are made since base crit should be irrelevant to the ratio of HPS/Stat vs MP5/Stat. I've redone the calculations with my own gear (7107 Intellect, 10042 SP, no buffs other than WoA, 1268 Haste, 1490 Crit, Burning meta):
+100 Haste is 161.46 HPS. +100 Crit is 108.25 HPS. The difference is 53.21 HPS. The MP5 difference is 40.48 MP5, which at 7107 INT translates to 57-58 Spirit. Both Crit and Haste translate to > 1 HPS per secondary stat, so ~57.5 spirit reforged is necessarily better than 53.21 HPS.

I want to add that while Haste looks better than Crit here by quite a bit throughput-wise, this represents nearly best case scenario since the timeframe is 1 "rotation" while if projected into infinity in a normal fight Haste runs into the problem that 2 out of the 3 spells are bound by cooldowns. This is why my first reaction when reading Jessamy's comment ("Every time I crunch the numbers I find that haste, crit, and mastery all improve healing throughput by similar amounts.") is that it's true, but Crit has the additional benefit of being a regen stat and via the magic of reforging you can convert regen into further throughput.

This is true. 7k Intellect is the number for the 31.58->45 Spirit conversion, while at 8k Intellect the conversion is to 42 Spirit. Considering that the HPS difference is 23.4 in the example, however, I think the amount of Intellect needed to totally inflate the value of Spirit to where it's equivalent is unattainable. In the interest of being complete I input 12k Intellect, into my spreadsheet, and the Spirit equivalent was 35 Spirit, in exchange for 29.3 HPS.

I think you arrived at your conclusion in what is somewhat of a roundabout way. Haste translates directly into extra casts, where 1% extra haste will give you 1 extra cast every 100. It is somewhat muddied by the fact that Haste is multiplicative, but we can agree that it's slightly less than 100 casts. If we say that Crit takes over 100 casts to do so, that is right in line with my consideration that for TC, Haste > Crit >>> Mastery.

I think you are missing the point. Mana at the end of a fight doesn't matter, so looking at a sterile which one gives you more casts over a fight mentality isn't of much use. If you have less than 20 seconds at 916 breakpoint or less than 12.4 seconds at the 2005 breakpoint, crit is going to benefit you more than haste. You will get more mana back over those casts giving you more mana in the thick of the fight. I think that is of much more importance than if you are going to get an extra casts every 80, 90, 100, etc over the course of a fight.

Futtër writes on the german forums on the topic of focused insights mechanics:

Since the setup is not easy to obtain (being goblin, having 2100 static haste and access to a heroic trinket), I suspect this will be more of a theoretical excercise.

Haste does not directly reduce cast time; it increases your character's spellcasting speed. Having X% haste enables you to cast 100+X spells in the timespan that it normally takes you to cast 100 spells.

New_cast_time = Old_cast_time / (1 + Haste)

Your base spell cast speed was 2.45 sec for CH before you hit the 2899 haste rating mark, which reduced your cast time to 2 secs. Thus if you remove Bottled Wishes, then your haste drops from original 22.637% to 19.06%. Using the formula for calculating new cast time (new cast time = old cast time / (1+ %haste/100)) you get 2.06 secs, where your old cast time is 2.45 secs so the math is fine.

Guys I am at a lost here. How important is Elemental Precision in the ele tree if you pick up Telluric Currents in the resto tree?

Adding on to this.

There is only a 5% difference between lavawalker and ghost wolf (unglyphed). In DS Heroics, Lavawalker is probably better than instant ghost wolf.

Considering our current 4 piece gives SWG a 20% haste + 5 seconds buff only increases the benefit of lava walker. It is probably better to drop the 2 points in instant ghost wolf for something else.

Since well handled DS heroics allow for a decent amount of time to be spent casting lightning bolt's you might as well drop it into the hit talents to improve mana regen.

I think there is another important aspect of the choice to take Elemental Precision. I think most shamans take Acuity regardless of your opinions on crit, so to get 2/3 in Ele Pres you still need to sink another 4 points into the Elemental tree. I don't know about you but my points are pretty tight. When I used to run Ele Pres I had to drop points in Elemental Weapons and Improved shields. I just don't think losing 40% on Earthliving and 15% on Water and Earth Shields is a very effective trade off. Sure you are going to miss 17 out of 100 casts, but I think in most of the situations where you can spam LB that number isn't going to make or break your mana pool. If you can squeeze the bare minimum out of the Resto tree you only have to give up one of those two, but I run a pretty tight spec with 33 in Resto, and personally I wish I had 1 more point in Resto to top off Ancestral Resolve...I just can't spare the point. Your thoughts may differ though, but I think in any situation you can't take both Elemental Weapons and Improved Shields if you elect to take Elemental Precision.

You're right Chuupag, you do have to throw a total of 7 points into the Ele tree in order to get the 2/3 Ele Prec. However, I'll refer you to the first page of this discussion where each talent is listed out. There you can see the opinion on BotE is optional for various reasons. It's value is determined by all sorts of variables like 10 man vs 25 man and tank healing vs raid healing. I personally found it worthwhile as a 10 man snipe healer fighting with a disc priest and resto druid to drop these 2 points in BotE to fuel my TC build.

8/2/31 Talent Calculator - World of Warcraft is what i run with. Is it perfect? Nope. It's what I've played with and found to be a good balance for me in my raiding situation and style of healing.

I'd point out that the way I spec'd I still do take the 2/2 elem weapons in the enh tree, and that the Improved Shields is no longer as powerful as you might remember it ever since they changed Water Shield with Resurgence. The mana gained from being attacked does go up with Imp Shields spec'd but resurgence does not.

If you look at this log of my Spine HM kill last night: World of Logs - Real Time Raid Analysis I find this to be one of the most annoying fights in relation to water shield and it's constantly being knocked off and having to be refreshed. As such, i think it's a great fight to look at the advantages or lack thereof associated with Improved shields. I gained 60k mana from my shield being broken. 140k from healing criticals, 296k from TC, 101k from replenishment etc etc. 2 talent points to get 15% more mana from my shields (60k) vs ensuring my lightning bolts hit and hit harder just doesn't seem worthwhile to me. Yes, it also increases earth shield healing done by 15%, though earth shield healing done is generally lower on the totem pole. 3% of my throughput on the spine fight, 2% on Deathwing HM.

End of the day, you have to look at your comp and playstyle and judge accordingly. 10 vs 25, raid vs tank, who you're healing with, what fight you're on, what stats you prefer and where your gear's at.

Can you hit the spell hit cap? Time for napkin math!

I've got an alliance shaman alt but I'm a bad so I had to go to Spell hit - Wowpedia - Your wiki guide to the World of Warcraft to find out with Heroic Presence - Spell - World of Warcraft you still need 1639.14 (rounded to 1640 from here) hit rating to reach the cap. If you only take 1/3 Elemental Precision - Spell - World of Warcraft, you're going to need 4920 spirit to be at 17% spell hit as a space goat.

Is 4920 spirit reachable? I'm not in full bis gear and not all my gear has spirit on it pre-reforging and I'm at 3068 spirit before my HotU is proc'd and 3948 after it's at 10 stacks. That's still almost 1000 spirit shy (about 3% off the hit cap still) and honestly, I wouldn't stack another 1000 spirit over haste/crit/mastery.

TLDR: No, you can't reasonably hit the spell hit cap as a space goat while only taking 1 point in elemental precision. You would need to stack 4920 spirit, which although may be possible I would not advise it.