Sorry, that you had bad experience with VEX GPS.
Despite being a very expensive sensor, it left many people disappointed and, probably, was rushed into production for the purpose of supporting IFI patent application.
There are still many unanswered questions regarding its internal algorithms and how well they are suited to provide navigation solution for a moving robot.
Since you cannot return it back to VEX or exchange it for inertial sensor, I would try to use it strictly as way to verify/reset robot’s absolute location every 15-20 seconds (while standing motionless for 1-2 sec) during programming skill runs and rely on odometry for the rest of your route. However, before GPS was released, resetting absolute robot location was traditionally done by touching field barrier, field element, or reading IR line sensor while driving over floor tape lines and it worked great for many teams.
So, the best tip would be to invest you time to learn about PID, odometry, path planning and tracking algorithms like Pure Pursuit, for all of which there are many pointers and tips, if you search this forum.
And, in case if anyone from VEX/IFI is reading this, in my opinion, the only way to turn VEX GPS from an ultimate lemon and failed product into something useful for VRC and student education in general, would be to open source its firmware and allow any team to modify it and use sensor’s camera and processing power in any way they see fit in VRC context.