So what happens when a griefer guild shows up and slaughters all the wolves and bears in your forest? How do prevent this or can they even?
What prevents real-life griefers from doing so? For one thing it's a massive task, not easily accomplished by twenty or thirty people (even if they're expert hunters). Another thing is that they would attract attention from local police, then armed militia, and in most places armed militia is more combat-effective than any civilian organization. Just implement that in-game. And if a faction does manage to overcome armed opposition and the sheer amplitude and execute such an amazing feat, then I want to see the ecology ruin and the local economy plunge. Just make it easy to regenerate scenario procedurally (the nethack approach -- things may die, but then you just play again with new things). If the players managed to ruin the whole world, why not have creator gods come up with a new one? Why not challenge these players to destroy the new one too, patching the game to be less and less exploitable -- wouldn't it be much more rewarding to the griefer guild to be known as destroyers of worlds than "those guys who narf n00bs in the town"? Hire some professional writers to come up with convincing explanations, the possibilities are endless.
I know, most gamers are power trippers and your level 99 "hero" needs to be the Strongest Creature on Earth and single-handled trample entire societies and gamers would oppose to be less powerful than guards. I for one wouldn't mind less powerful characters in a more immersive world. Hell, I bet I'd feel more powerful if I could somehow affect the world, however slightly.
If it is so darn "not difficult", why haven't you written your own game and have a few hundred thousand subscribers already?
Er, because it takes hundreds of people and thousands of dollars to put a 3D MMO online? I did experiment with roguelikes; I hacked a bare-bones ecosystem in a weekend in Ruby, and now I'm (leisurely) playing with fractal terrain in Scheme. By now I'm convinced a simulation-centric (as opposed to stats-growing--centric) MMORPG is feasible; it just wasn't tried yet.
Never trust an operating system.