I agree with you guys, it can be all those things. The more I learn the magic system the more I appreciate it's flexibility - really fantastic! I've been too rigid in my thinking - I guess D&D has me expecting very defined spellcasting 
It's not just spell casting, it is everything.
For instance, you can even resolve a major conflict with a simple or extended contest if the players agree... Lets say you have assembled a somewhat large team for a big battle, you have all the PCs (4 players), 4 white council wardens, and 2 summer court folks, up against 12 demons and the sorcerer that summoned them.
Rather than taking a few hours to fight the whole thing out, roll an extended contest. Each player rolls using their strongest offensive skills (feel free to invoke aspects), and they also roll for the NPC groups on their side using an average skill level for the groups. All this is summed for the players side of the battle. You also keep track of each player and groups individual result for later.
You then roll proportionately the same number of rolls for the demons, 1 using the sorcerers best skill and then 6 using the average demon skill (the demons have 12 on their side vs. 10 on the players, so they get one extra roll). Then compare the sums, and determine which side wins.
The more shifts the winning side gets the more people the losing side loses (not always dead, but out of the fight for now). Whoever rolled the worst on the losing side is the first one taken out, second worst, second taken out, etc. (those in "groups" just lose one member per loss until paid up). From a story aspect the people that rolled the best on the winning side are the ones that took out the ones that rolled worst on the losing side, allowing the players and GM to put together an interesting narrative despite the simplistic combat. Once the dust is settled continue this process until there is a winner, or until everyone agrees it is time to go back to "normal" combat because the numbers are small enough. One caveat... At least half of the players should make it to the end fight no matter how bad they roll (possibly by giving them a chance to help each other recover that the NPCs don't get), and the main antagonist (the sorcerer in this case) should also make it to the end no matter what.
If you want to do this even faster, just reduce the numbers small enough for the final battle based on just the first contest, no extended needed. So if the sorcerer wins by 5 shifts then the PCs were beaten up pretty bad, the are down to 2 PCs and one Warden vs. 4 demons and the sorcerer.
Or, if you want to do is super fast just chose how big the "Main" conflict should be, 4 PCs vs. Sorcerer and 4 demons perhaps, and just assume the other people are off fighting each other and how that turns out will be decided by how the PCs do. The downside to this method compared to the other two is that if any of the NPCs on the PCs side are important to them, like maybe one or two are actual faces, you don't have a chance to narrate their fate directly.
