There's a thing called "time-to-market" and sometimes, having a product delivered sooner rather than later is a better option. I'm already starting to use SuperPowers (which includes 3D) support via ThreeJS. The point is, when people start looking for alternatives and start to get comfortable with the other solutions, they may not have the time to "go back" to Construct 3 for newer games. They may use C2 only to maintain older games (which i'm doing now).
C2 is a doorway for so called "non-programmers" to get into game development. But it's a matter of time before said "non-programmer" acquire enough programming skills to get into the more sophisticated tools. It takes less time to learn a new language, get to an intermediate skill level and then use Unity than it is to wait for C3 (as an example).
I would suggest C3 development and roll-out be in phases -- if even possible. The community would likely be more supportive of a program that is forever in beta than to have something 2 years later that may turn out to be just as common as the next tool (IMO).
I don't need a better iPhone, I need to be better.