Open-Vocabulary 3D Object Detection with Co-Distillation Discovery and Dual Guidance Robust Training