It's almost impossible at this stage to talk about which class can generate the most threat or has the most survivability at level 80. There are no level 80 characters in beta, and we haven't done our own testing yet. We want to come up with mechanics we like, then we get the numbers in good shape. We don't talk about it because it's very useful when someone can point out a potential problem, particularly if it's one we hadn't thought of.
In the world of warcraft, there are four tanks. They all are intended to tank 5-player, 10-player and 25-player instances. They all have their specialties and the Warrior specialty will probably remain as the best tank for single, hard-hitting bosses. But if you only have a death knight for that encounter, or you bring a Warrior to a fight with a bunch of adds, you'll still be able to get purps. This is a slight change in philosophy for us, but one we feel is necessary in a world with 10 classes and several specs getting a boost to raid viability.
And also we acknowledge that hit and expertise are great threat stats, and expertise can offer a little mitigation to boot. The point is that putting strength on tanking gear solves a lot of problems in the game — it can improve DPS and threat without us having to worry about whether plate-wearing tanks are already capped in some other stat. Defense means something different for Warriors and Paladins than it does for Death Knights, and unless we build different gear for each class we can't count on defense as always being desirable above everything else. The last thing we want is for some classes to feel that they don't have access to the gear to do their jobs properly.
We get to know that there is a tradition in Burning Crusade of a prot Warrior MT with perhaps another prot Warrior or a Paladin as OT. If we do our jobs right, there will be some gouprs that run feral MTs with unholy DK OTs in Lich King raids.
As to Warrior, AE tanking is hard. In particular this tends to mean that pugs would rather have a Paladin tank just for the consistence. We think we can make it easier for Warriors to AE tank, especially in 5-player instances, without displacing the Paladin as the best AE tank. At the moment we are considering increasing Thunder Clap to 5 targets. We'll see how that feels.
It wasn't fun to make the old Shield Block. But we think we have the abilities to make the new one fun. Numbers are the easiest thing to tweak if that's all that's called for. So we tend to focus on mechanics at this stage in development. Once we like the mechanics, we can massage the numbers.'
For the most part, Warriors need to be better AE tanks without eclipsing Paladins. Thunder Clap is a good place to address that problem. If Shockwave becomes the ultimate tanking ability than we're concerned nobody would want to run a 5-player dungeon without it. That's not the goal. We don't want to hand out Consecrate to every tank, but we want you to be able to tank groups better.
Yet they haven't a clear notion how threat handling / management will happen at level 80. They want to keep each tank with its own specialty but improve overall viability and want to make it perfectly possible and viable to raid without any one definite cookie-cutter tank spec. What's more, they want to look at any formerly "must have" talents and integrate most of these into core mechanics to foster more varied spec flavors instead of one single cookie-cutter approach.