OP, the stat priorities have been listed, however, don't try to aim for specific arbitrary values. Build towards BiS lists as best you can - those have been tested thoroughly and run through solid equations. Then for the pieces you aren't able to obtain you can still follow the priority and be good enough. Just be mindful of target GCDs for your job to make the rotation sane enough.

Regarding Dragonsight if you can manually resolve tether seamlessly enough on a given party member that's a better case, but you can have macros for specific party slots just to reduce the number of different inputs as a compromise.