" There's little attempt to really explore the possibilities of multi-player.'
On the contrary, companies like Zynga have explored and researched the possibilities of social interaction a hundred times more that most game designers, tweaking all the tiny elements to an optimum range that works on the hairless apes on Facebook - that is then copied to all the games.
All these shallow elements that you mention - they work. They keep the most amount of players coming back. The recovery rate of "energy levels" and waiting time of various in-game activities are tweaked to have the most people log in back to the game. The allowed interactions between you and your "neighbors" are tuned to drive you to interact with as many of them as possible and have players motivate each other to stay in game.
Deep player-to-player interaction and tough challenges ? Meaningful interaction between Johnny-avatar and Jimmy-avatar with actual choices requires non-zero effort and has a chance of conflict, and has less cases where Johnny pokes Jimmy out-of-game saying 'log in now and assist me on Genericville!'. So they are deliberately filtered out of the design because clearly they bring poorer results.
If you want the truth, don't listen to what people say about their preferences, but look at what they do. No matter what features and gameplay people say that they want, they have shown with their mouse clicks what game features they are actually playing, and these Facebook games have proven that (most) people actually want a stupid button that gives out shiny reward-like emotions at an optimal interval.
After all these manipulations, more people come back to Zynga games than they come back to "proper, good, serious, deep" games. I'm not saying that this is good, but that's how it is in real life - no matter what gamers or critics or designers might say, in practice for every person that would even consider playing FPS or RTS there are ten that prefer Farmville.
Ergo, if a game design theory says that Farmville is inferior to a good FPS or a good RTS or a good RPG, then the theory is simply flawed and false, as it doesn't match what we are seeing in real life. And it's useless to argue about how it should be - just as gravity makes balls roll downwards, our brain reward chemistry makes some "social" gameplay elements more effective than others.