Start a new project with technologies you are familiar with. Try new technologies on projects you are familiar with. Don't start a new project with new technologies.
i would say try new technologies on problems you are familiar with and focus on the problem space itself otherwise. that is independent of if this is a new project or not. if my goal is to learn something new i try to find a project in a problem space i am very familiar with, where i know the pitfalls and rough edges that make things harder than they should be. this gives me a much better understanding of whatever that new technology is about and if i really want to buy into it...
i.e i learned some programming languages by implementing a simple binary Usenet reader/client. i know pretty well by now what it encompasses to build that thing why it is harder to do with some paradigms then others and immediately see value whenever something turns out to be more elegant then previously. and if the task gets unpleasant i can usually state if that is because of the task or the technology or in this case the language i am using....
i.e i learned some programming languages by implementing a simple binary Usenet reader/client. i know pretty well by now what it encompasses to build that thing why it is harder to do with some paradigms then others and immediately see value whenever something turns out to be more elegant then previously. and if the task gets unpleasant i can usually state if that is because of the task or the technology or in this case the language i am using....