Yes, quality of games is a difficult point to discuss.

And I can understand different positions there. I made games in the past myself as a student. And I did not get much money per game from the publisher. So I had to make small games quite fast. That way I decided to go the fastest route each time. I programmed in a very good programming language with a powerful library at that time and I made 2d-games only. So I could sometimes finish a game on an extended weekend (Friday to Sunday).

And because of this I think it still works that way, many small lone developers have to make several games per year just to survive, otherwise it is a pure hobby. And it will be difficult to start coding your own tools and libraries for each game. They also have to get to as much platforms as possible to earn something.

I have read that Torque2D attracts a lot of people who really earn money with games, Shiva, Flash and Unity also does out of the reasons I mentioned above.

I am sure you can also earn a lot of money with a really good looking and bigger indie project, but you have to invest a lot of money in front and this will be harder.

I am dreaming myself of making a great looking shooter and I probably will make one. But it is a long time project and might never become very successful. And in this case I would probably rather use UDK even when Unity has a more easy workflow. But my tests show that I can get better looking results even render faster there.

I can read such discussions each week in other forums. T3D and C4 users claim their engine choice is the best just because they have the source code to adapt whatever they like. C4 users often write they already know Torque and C4 code is just better organized. Both suffer from worse tools compared to Unity and UDK. Unity and UDK users love their tools just because of the pure power they have at their hands. They might bounce against some walls but they help each other to solve it or just work around it. In the end it works for most of them.