My big question now is: What could be useful about C2 technologies if it is at the end only produces poor performing mobile games / apps?
I think the best way to give you an idea of the performance is to go to the showcase of each engine and download or buy the games done with it (free or paid).
The ones I've tried shows rapidly some results, in the order of "performance" on mobile :
- native apps <img src="{SMILIES_PATH}/icon_e_smile.gif" alt=":)" title="Smile"> yes it is probably the best way to go,
- Haxe/OpenFL/AS3/Air (a game like Rune Raiders or Half Inch Heist shows great performance),
- Unity3D : needs a good smartphone to run games in 3D,
- HTML5/JS games (C2, Canvas2D/WebGL technology : LinkPop, the Flag Bearer - search for complete creations here on the forum).
I noticed the Air platform is very satisfying and that's why we can say "Flash is not dead". All the apps using the Haxe language ("One language to rule them all") can export in a variety of formats, from native to Air to HTML/JS to almost whatever you want (because Haxe use one language derived from AS3 and converts the code to other languages before compiling it fro the expected target). But it needs to maintain a lot of librairies for the developpers of the tools like C2, Stencyl and so on.
Another FW that use Haxe with export to HTML/JS and Flash : https://github.com/aduros/flambe, there is the Phaser FW to code directly in HTML5/JS.
Again, from my point of view, C2 is great because of its interface and good performance mainly for desktop/browser games. If you wanna go mobile NOW and are ok for coding in Haxe or targeting Air or Flash, you should try something else (Haxe supports a lot of gaming FW like Flixel, Flashpunk and almost FW can be added to Haxe).
I repeat actually my concern is to rapid prototyping games, for the browser or the desktop, and not put my hands in the code too much (ie, make a base platformer in 1 day rather than in 5 days ^^).