While I will agree that 3 mana to cast lower land is a bit ridiculous, this is a simple balancing issue. Honestly, I have never seen the AI actually build choke point cities which require me to uses these spells to get around the choke point. In the end, I like the idea of being able to blow a hole in a mountain wall, as it reminds me of the final battle in Dune.

This is a direct consequence of the removal of essence. Before, big game changing spells like terraforming could cost essence, of which you only had a finite amount (and it had a lot of other uses). Now it just costs mana, which is easily gained by hoarding it a few turns.

It's nice that these spells are in the game, but casting them should be a little less common, and be way up highin the spellbook. If I'm separated from my enemy by some water, and he casts raise land to create a bridge, it should be a "holy cr*p, he actually did that" moment, not the "well, that was to be expected" moment it is now...

Raise/lower land used to be a single tile in 1.09 - now they both affect 9 tiles and the mana cost is cheap. I usually lower mountains in my zoc to provide direct routes for caravan road building (helps improve logistics) and raise land around cities on the coast to expand tile building or give caravans the shortest route possible.

Regarding teleport, I'm finding it useful to plug any gaps behind my lines, never from faction AI, only from mobs. I created a thread awhile ago with some thoughts on how to nerf teleport by requiring a solid development path. The ability to teleport an entire army stack to any friendly tile should be the uber end-game benefit (only thing better would be to teleport to any tile, anywhere), offset with some kind of appropriate end-game counter. Each step along the development path should provide an improved benefit and an appropriate counter. There's been numerous other good thread ideas too. I don't know how in the world Brad can code the AI to deal with teleport as-is: in the span of a single turn the threat meter can go from 0 to 100. With normal logistical rules the AI should see a threat moving towards one of their cities. Teleporting to any tile in your ZOC and then blitzkrieging through enemy lines must throw the AI fits as it would an human opponent. Maybe there should be movement penalties when in enemy land? Maybe when initially moving into an enemy ZOC tile the stack or unit's turn should end? Maybe there should be a global enchantment spell that keeps a permanent storm cloud over enemy stacks to lower movement to 1 or something? We need something, or a combination of things, to allow the AI to adjust and rally the defenses. Not to mention multiplayer balance.

But I too a m finding the urge to create my own house rules to dictate how I should use these types of spells.

Completely disagree.

I really love the idea behind being able to manipulate the land too, and I've used it a few times on different maps to gain access to resources that the AI intelligently (or randomly) blocked me from, but I felt a little guilty using it.  I agree though, this could be solved with some balancing of the mana cost or even a timer on the spell to make it not usable as often.  For me it was kinda like having a cheat book while playing Myst... you could sit there, figure out the riddle/puzzle/etc. or you could just go to the book and find the solution.  Removing a mountain feels a lot like taking the easy way out

Also, on a different but related note, shouldn't the AI be kinda pissed that you build a settlement right by their food source and only have to wait a few turns before your influence gives it to you? I can tell you that I have gone to war with factions for less than that.

as far as i know, we  cannot cast raise/lower land in an enemies ZOC, correct?

I'm just waiting for the tactical battle changes where individual units have a turn speed and alternate based on that. Just that will make the tactical battles and the overall game so much better. Unit inititive I know is a big thing, but it was stated it will be done and being a big task, it will have huge rewards imo.