I've got lots of things to say about the minecraft modding community from end to end having done some stuff in the background of both sides. From development, to server management, to being an end user of all this.
I think that the only hope for a good API is to actually have an API. Bukkit was the correct way to go from the start, forge is not an API.
If anyone want's to know funny stories about bukit, forge, minecraft modding, and all you can get in contact with me and I'd be glad to talk about it all.
However, most of the more interesting mods still could not possibly exist for Bukkit, due to it's server side status, so I can't see any MS API faring any better than Bukkit
"Sponge is a combination of a new API (based off of Spout/Flow’s APIs) implemented on top of Forge, with assistance from other parts of Minecraft’s modding community (Glowstone, Cauldron, Spout, etc.).
It will be both a server and client API, and its target user base is pretty much anyone that wants to mod their game, including server owners. However, we may focus on the server-side portion first.
We invite any developer to help out."
https://www.spongepowered.org/
Development seems quite active on Github.
They are even making some of the VERY basic yet devastating problems that Bukkit made at the start. Again, there is more to go into than I can fit into the scope of this comment chain.
Bukkit failed because of copyright laws. Despite this, people still use bukkit. It was powerful, easy, and had an intuitive API. It made programming seem fun to me.
A proper API done client side would look like bukkit, but have hooks into almost anything you could need. Look at GMod for example.
I tried MineTest but it really (!) wasn't as easy to get started modding it as with Forge-Gradle. You can see my little mod at https://github.com/voltagex/minecraft-rebridge. I want to go back to i, one day.
Then again, part of that is that I'm a clumsy coder, and part of that is Forge's "support" channels.