Writing a procedural world game is a good test of an engine. Will it let me do what I want, or am I basically making a mod of what the engine author wants? And 7DFPS is such a great time to do put them to the test. "Will this thing let me innovate?"
2012 I learned Ogre and made a procedural island. Painful to get started, but easy to do the procedural bit once I was.
2013 I learned Unity and it was very easy to start. The procedural bit was annoying, but easy enough that it was still faster and easier than ogre over all.
2014 I got Unreal Engine 4. It took me a day of forums and IRC to make a procedural surface interact with physics. Another few hours of forums and videos and I still can't find a way to texture it, and I give up. It might be a really good engine, I might be just missing the point or rejecting a learning curve, but right now it is getting in my way more than it is helping and I don't have enough energy left after a day job to care.
I am quite sad to miss one of these events, but I guess every experiment has a "failed" condition, and I was bound to one day.