What are the limitations on survey logic complexity? Can we implement nested conditional logic with multiple dependencies? Can we create logic that references answers from different survey sections or repeat question groups?
Yes, we can do all these things.
If you can put a particular situation, then we can discuss that in a better way.