First off, there's the semantics. What is a social game? Is it something running in Facebook? If so, you're not going to see it evolve too much too soon. That's because of the way the platform incentives are structured, the mindset of the average person who is on Facebook at the moment (more often than not at work, multitasking, able to devote only infrequent, short bursts of attention) and the inherent technical limitations of the Flash platform. If you ever saw the traffic graphs of a successful game, you'd understand immediately why casual is winning. It has a high bounce rate and peaks on weekday mornings.
If you define "social gaming" more broadly, which I think you should, then World of Warcraft, League of Legends, and many other very innovative games are social games. In that case you're already seeing social games evolve. My little cousins play Call of Duty over Xbox live with each other from their homes 50 miles apart every day. That's social gaming if you ask me, and they at least think it's fun.
There's certainly no accounting for taste. It's just plain snobbish to say the games you like are "universally recognized as fun" and the games you don't like aren't. I'd rather jam a ballpoint pen into my eye than play an asynchronous game of Scrabble. You'd sooner catch me playing Farmville.
For the purposes of a thread about Zynga, it's reasonable to assume a social game in this context is, indeed, a Facebook-flavored game.
Realtime multiplayer has enjoyed a lot of evolution since you had to bind curses and status reports to your F-keys in Quake. Definitely share your enthusiasm for where it is going right now.
Asynch casual multiplayer is in a dreadful rut. The big beef I have with it is that so little of it is skill based that you can't enjoy improvement over time, except with more and more swipes of the credit card (edit: or pointless grinding). I don't think the games I like are universally recognized as fun. What I said, if you'll read my post, is that asynch casual games are relying more on ill-concealed compulsion loops than on anything an observer would look at and think "boy, that looks fun to play."
Until asynch casual multiplayer, as your semantic nitpicking requires I call it, leverages gameplay that lets players feel themselves grow more skilled, we're just not talking about anything of lasting value. Personal growth is a huge component of enjoyable gameplay and you can't just add a row to a spreadsheet game to simulate that.
It's social, casual, skill based, and fun.
To answer a likely question, it's social be use I want to beat my friends scores each week
Scott Adams wrote a very good article about the illusion of skill this week. Skill is mostly the result of a compulsion loop. I play Scrabble, I improve, so I play more Scrabble. It's not much different than harvest crop, get money, build another crop.
If I play Scrabble more than you, I will probably beat you when you and I play Scrabble. If I play Farmville more than you, I will have a bigger and better farm. Both of those are generalities (it is of course quite possible to play both games sub-optimally, though easier with the former).
Also that's not the only thing anyone considers fun. A lot of people look at a virtual farm full of crops and chickens and barns and think "boy that looks fun to play". Millions. There's more to it than compulsion loops, they don't keep players around for very long.
It could use more depth but match three as a game style has some compelling play.
You seem to have a great vision for what social games could be. I cannot urge this enough - please go build that vision now. The industry needs innovation badly. There are huge profits to the person who can pull this off, and I hope that person is you.
But my message is this - Build Stuff People Want. Many people have fallen into this trap, myself included. I thought I could buck the trend, out-innovate Zynga, and make a truly fun game. But in the end I failed because I spent too much time in an ivory tower making the perfect game, and no time making an enjoyable product.
A great analogy is engineering brilliant software, but failing to build something people want to buy.
Inception was a success because Christopher Nolan is able to carefully balance the things that make a great movie with the things that make great ticket sales. It wasn't the best movie ever, and it wasn't the best selling movie ever. But a careful blend of both lead to success.
No one in social games today can do this. I want that to change.
To think that one can simply "made something good" is a fallacious line of thinking along the lines of "build it, and they will come". Fundamental understanding of what your audience wants is so critical.
I'm not saying Zynga makes good games. I'm not saying you should follow in Zynga's footsteps. I'm saying you need to understand what Zynga does right if you want to beat them. And if you think all Zynga does is "be evil" and "buy ads", you haven't been paying close enough attention.