(This is a redux of what I post regularly on this forum on the subject)
I'm not convinced an easy-3D-game-builder product is as realistic as you think, especially if trying to keep Construct 2's philosophy of "not a cookie cutter engine", and mainly trying to provide primitives from which to build a totally custom game from.
Fimbul touched on a few issues but a few others that I think would be an issue:
- 3D math is a significant jump in difficulty over 2D math. Most students touch on sin and cos before leaving school, and they're really handy to know for game design. When it comes to 3D math, how's your knowledge of quaternions and matrices?
- Designing 3D assets is a significant jump in difficulty over 2D assets. Most people can scrawl something in Paint or do some light pixel art, but in comparison how many people are competent at 3D modelling tools? Also a 2D layout editor is nice and easy to use, but custom 3D level design on a 2D screen can be a pain in the ass (at least in my pretty limited experience of designing half life levels way back)
- Competing with the visual quality of the top-end 3D games (think Crysis-grade and newer) is tough, but IMO it's realistic for indie developers to match commercial quality with 2D titles. (See "The Next Penelope"!)
I think a "Construct 3D" is the top example of "imagining feature requests to solve everything magically with no effort". Real-world solutions always have their caveats.