I was checking out a new GUI toolkit, and my first impression was "wow, this is a lot harder to use than my previous toolkit." But then I wondered, "is it harder because it is unfamiliar, or because there are genuine issues?"
What are some techniques for objectively evaluating new GUI toolkits?
I would look at how difficult it is and how long it takes you to write a certain feature. Then look at the total number lines/steps it took. And then finally look at the end result. Comparing those helps make it easier to decide, just write a small test program if you need to. Also the end result involves not only appearence but also platform dependency, support, and speed.