There are great suggestions in this thread! I like that you will only implement spells that the AI knows how to use, but I would hate to see magic, of all things in this game, to suffer for insufficient AI. Frankly, if a game called "Elemental: War of Magic" is released with lackluster magic, with largely straightforward and simplistic spells, it will be a big disappointment and I imagine it'd have a significant effect on its reception. I would rather see Elemental delayed until Winter than be released with lackluster magic, due to lacking AI. I mean, if the AI can't handle anything but the most straightforward spells, how are we going to mod in our own, nifty spells? We'd have to mod the AI to be able to understand them, and that does not sound like a trivial task.
One issue I have is that all your elemental DD spells like Hurl Boulder and Fireball are exactly the same, besides their damage type. (EDIT: I just noticed that the mana costs are slightly different, too) Some should be stronger than others. Some should do little damage but have significant secondary effects. Some should just be downright better or weaker than others (while other spells in the category with the weaker DD spells would make up for it). The fact that Lightning, Fireball, Hurl Boulder, etc all have the exact same effect when cast on a unit with no resistances is a Bad Thing. Same with Ice Bolt and Arcane Arrow.
There are so many ways some of those spells can be spiced up, and probably many ways that the AI could handle relatively well. Hurl Boulder has a lot of potential, for one: boulders roll and bounce! The boulder can hit the main target, and it could continue rolling, hurting units behind the target for diminishing damage. It'd be particularly awesome if we could choose which direction to throw the boulder (who says the boulder has to originate at the caster?) - I'd love to watch my caster launch a distant rock into an equally distant enemy, at a skewed angle.
In my opinion, not all of those schools even need single-target DD spells. I'd be quite happy with none in the Fire Spellbook; it could have single-target DoT spells, AoE DD spells, but could lack a single-target DD spell. Even if the spells are largely straightforward and simplistic, true diversity between the books would go a long way to making magic more fun.
This problem is reproduced in other areas as well. For example, there seem to be a lot of different ways of destroying resource tiles. Would it really be so wrong to limit abilities like that to one or two spellbooks? What is the impetus in researching new spellbooks, or new spells, that do pretty much the exact same thing as a spell I already have? Some ~repetition here and there is fine, but based on that chart it looks to be rampant (though, as always, perhaps within the game it isn't as bad as it seems like it'd be).
Also, Life vs. Death. There is zero difference besides the type of land that you revive. I realize that you're using Life and Death to provide all players with the basic, core set of spells that everyone will probably need. But it would be nice if there was a little diversity even there. Maybe mix up mana costs (so some spells might be more expensive in one than the other, and vice versa). Maybe add a few exclusive spells to each, or have some spells that accomplish more or less the same thing in Life and Death, but in different ways or with small differences.
Subtle things don't have to be complex things. Subtlety can be very simple, and still very evocative.
One other thing, somewhat related: please consider changing your current "N essence means you can maintain N enchantments." Please make some enchantments harder to maintain than others. If you don't, then the less effective ones will never be used if the better ones are available. Implementing this, as someone else in some other thread pointed out, would be extremely simple: just make some enchantments require more than 1 essence to maintain.
