Alright so I'll try to tackle this question for you:
First: Raw int does very little overall because most are well over the caps
Second: HINT is minimally impactful (Small boost to overall mana pool, 1 mana regen/25 hint and effects spell and dot shielding mod2s now)
Third: Mana regen stat only does something if you are under your mana regen cap
Four: An increase in mana pool only helps if you run out of mana, if you don't run out of mana (e.g. frequent delays between pools, med breaks before out of mana etc) it does nothing for you.
Spell damage: The formula for how this effects spells is as follows
Rich (BB code):
(Cast_Time + Recast_Time) / 7 * Spell_Damage
Truncate the damage (e.g. round down)
So lets look at what 30 spell damage would do when a magician casts Spear of Molten Shieldstone (Cast time:3.5, Recast time:9, base dmg: 26967 (before crits and focus and other mods are taken into account)
This adds 53 dmg to each cast of that spell, however this damage can't be critted or focused and is added as a flat damage addition to each cast after everything else is calculated for the spell. Or in other words, over the 12.5 sec cycle for that spell it adds a grand total of 4.24 dps.
Now lets get into what clairvoyance does. Clairvoyance is a 2% chance to return mana equal to your clairvoyance stat on a spell cast within a level range relative to your level (cant find the post describing specific level range at the moment) and likely over a minimum mana cost. Lets use the above spell as example as at 105 it is current and thus should be included and has a mana cost of 4487 at base so should exceed and min mana cost. So with that said lets break down the difference in clairvoyance. To do that I will look at the average return (so what you would expect if you cast the spell 100 times and averaged the return from those 100 spells) as the value on a per cast basis. The Clairvoyance difference is 39 between the two pieces so we will use that as the clairvoyance value here to understand the difference between using the two items.
Rich (BB code):
0.02*39=0.78 mana returned on average per cast
In order for this to equal 1 cast of the spell we are looking at we would need to cast the spell on average 5752.56 times before it would yield 1 additional cast of the spell.
Now into the real difference between the combinations:
1hnd+shield gives you the proc Sympathetic Alleviating Burst (in this case +330 mana return) on spells level 70+ that cost 450 mana or more .
The 2hndr gives you the proc Sympathetic Hammer of Magic (in this case a 7k DD proc) on spells 70+ that cost 10 or more mana.
Option 1:In order to regain enough mana to cast 1 more of our nuke previously discussed with option 1
Option 2: In order to proc damage equal to the base damage of our nuke (Functional equivalent of above to yield the same damage output) we would need to cast
so if we now look at how much base damage in casts we lose by using 1hnd+shield before crits, foci, modifiers etc are applied lets look at 14 casts (since we can't cast a spell 13.6 times in the real world):
Option 1: 26967 base damage
Option 2: 98000 base damage
Difference: 71033 damage
Now assuming no crits, foci, modifiers etc. casting the spell 14 times would take 175 seconds
The loss in DPS going from Option 2 to Option 1 is ~406 dps (in reality way more when crits/foci/etc are all factored in).
Regarding the total difference in mana pool you would get ~11.5% of a nuke in gain of total mana pool which would only be relevant if you ever fully hit oom and as the above example highlights would still not make a difference in the overall decisions made due to the increase in raw damage done from the 2hnd proc.
I hope that summary makes it clear the following:
1) How irrelevant some stats are and why they should never be considered a primary component in the equation of gear selection
2) What the clear gear choice for a mage looking to maximize their DPS should be
