Things that didn't surprise me
Specialist being the highest-rated non-ally: This is just a great card. It's rare that you wouldn't want it.
Town having the lowest variance and the highest minimum: This is the sort of card that you can tell right away is good. It's useful for $ in the beginning and then becomes an important engine piece later on.
Forts having the highest variance of the non-allies: I suspect there will continue to be a lot of arguments about how good Tent is and when to go for it.
Swap having such a high variance: This seems like its usefulness depends heavily on the kingdom. I gave it a 0, not because it's completely useless but because I was only doing an inter-set comparison and this definitely seemed like the weakest card relative to its cost. It's basically a cross between Advance and Transmogrify, but it lacks the main thing that makes those two useful: the fact that they're "always there when you need them". With Swap you have to collide with a good target when you play it, which can be difficult. If you're relying on Swap to save your engine in case it duds, it seems like it would be better just to buy more villages, especially since villages are usually cheaper than Swap.
Underling having the second-lowest variance: Yeah, I mean it's okay. What else is there to say, really?
Things that did surprise me
Clashes having such a high variance: They had a high overall rating, which I agree with. But I guess a bunch of people disagreed, though?
Royal Galley doing so well: I don't know, I haven't been that impressed with this. I feel the fact that you have to wait till next turn to play the card again takes a lot of wind out of your engine. That could just be psychological, though. I acknowledge that getting draw at the start of your turn can be really really good, but having to set the card aside can also make it miss a shuffle.