They do inherit your +hit, however, they only cast frost or fire spells, which means they are going to be missing 3% hit. I also assume based on my own WWS that they do not benefit from Heroic Presence.

That doesn't seem to fit. I have 11.89% hit from gear, 3/3 AF, 3/3 EP. That means with a Spriest or Moonkin (which my raid always has) I'd have 20.89% hit. If Mirror Images inherit +hit talents but cast only Frostbolt/Fire Blast, they should still be over the hit cap at 17.89% hit, but they still occasionally miss. Either they're not getting +hit talents or they're not getting +hit from my gear.

From a WWS parse of Patchwerk on Tuesday (thereby eliminating any possibility that the spriest or moonkin may have died): Wow Web Stats

My Mirror Images missed 1 out of 36 Frostbolts and 1 out of 20 Fire Blasts.

This. They get your +hit from gear and from raid debuffs, but not from buffs (Draenei Aura) or talents.

Looking at the hit points of bosses in Ulduar, it seems like fights are going to get much longer, which would make mage armor more benefitial, however there will likely be times when you don't need extra mana regen and molten armor would be more benefitial.

Even though Molten Armor glyph provides the best DPS, I am not convinced going forward into the future that it is going to be the most benefitial glyph if you are constantly changing armors.

This leads me to believe that Arcane Power Glyph is the best of both worlds, allows you to choose which type of armor without sacrificing your glyph.

Of course the super hardcore will probably dual spec two arcane builds, which might not be such a bad idea as well, glyphing mage armor and not getting meditation, allowing for full protection on your casts and channels due to pushback.

That's the point. As you said, if you can last two minutes, you can last forever. It makes more sense to settle for a lower DPS rotation on occasion than it does to give up 5% crit.

For as long as DPM tradeoffs remain as low as they are now, this situation is unlikely to change. Given the current direction of arcane I doubt that DPM tradeoffs will vary a lot in the future.

Does this type of scenario put fireball closer to arcane or even ahead?

If it is close is it also safe to assume that the gear upgrades which favor fireball scaling vs. arcane scaling will put fireball over the top for good?

I think you can predict more than you're willing to admit though. Wouldn't it be more logical to assume the scaling will reflect 200->213->226 rather than some drastic new path? In a void, you're correct, but we have years worth of information to draw conclusions from. An educated guess would suggest that a mage in Ulduar gear will perform better with a Fire spec than an Arcane spec. Is it possible that the assumption is wrong? Of course (and I hope it is), but it seems silly to claim that it's impossible to predict.

It is predictable in a sense.

Any item with spirit will have two main stats. One spellpower the other Crit/Hit/Haste/MP5

Almost every top-tier item will have three stats, exception is listed above.

There were a few other exceptions, like the Wanton pants, that stacked more spellpower. Itemization used to be that you could shop around and focus on one or two areas. The homogenizing of gear has almost destroyed our ability to do that anymore.

There is a distinct "budget" for every item, some seem a hair low, and some seem over. Anything with sockets seemed to be budgeted with the average available gem, and could easily become better than the same iLevel gear with zero sockets.

With all of the new trends, we have to dismiss any budgeting mechanisms from tBC or earlier.

Since arcane has one of the lowest +hit requirements, it will indeed have a harder time trying to stack Haste / SP without soaring over the hit cap.

I'd agree that you can't look to before TBC; the "four years" that I sited was relative. The point is that itemization and fight mechanics didn't drastically change between AQ40 and Naxx (though itemization has obviously changed a great deal if you compare them to now).

Regardless, your first two lines are perfectly accurate and that is my point. We know all the variables, thus predicting the gear we will get is not difficult. I agree with Manly that it can be silly to deal in such absolute terms about something we have not seen, but I also think it is fairer than he seems to think. We could all draft up ideas for the stats we feel will be on the Ulduar 25-man gear and none of us would be very far off.

Manly is wary to speak of 'how things will be' both because he's absolutely right in that we can't predict it at all with any degree of certainty, and because what he says tends to get taken as true by a lot of people who lurk these forums (rightly so, a large majority of the time he is right), and to maintain that credibility it behooves him to not partake in musing about how things might be when there's so little information at hand.