The question is actually vague to begin with.
Selling a game to a publisher ? Selling a game over on Steam ? Selling games on the mobiles app markets ?
All the three above have been done by users on those forums using Construct 2.
The software in itself is not the limit. It doesn't matter what engine you are using in the end, if in the end you manage in making a product that is worth it.
C2 is not a scripting environment. "True coders" need to either embrace its workflow, or find their way to another engine perhaps more oriented towards writing hard code as they are used to. It doesn't make C2 less capable.
Hard coding is almost not required with Construct since its event system is so versatile and capable, but again, the user has to put some time into learning how to use it.
The SDK still allows for hard coding if really required, but most "pros" will try to dive in that way instead of learning the event system, which is a wrong way to go about C2.
I've worked with engineers who were able to take on Construct in a question of weeks.
And if you visit those forums you can still see users who have been members for years and are still struggling with basic concept.
If you already have basic programming notions, that should help you get into C2 quicker, but you have to learn its ways.
As for selling games made with it, it depends on the game in the end, not the engine.
Games have been made and sold. You can find some in the showcase. You can find quite a lot of games made (and possibly sold) in the "Completed creations" forum as well.
What are the stats ? Impossible to tell. Not all C2 users are aiming to sell games they make.
Some hobbyist are perfectly happy releasing free games.
Some people are still learning and making games for them.
Some people use C2 in their daily job.