Level optimization

[Zemanzo]

Loading the level was heckin' slow, now it's way faster!

Changes in this PR

• .glb support for models.
• Support for models using quantization techniques (using DRACOLoader)
  • This required an extra bit of configuration in webpack
• Several compression techniques resulting in a 56.72% size reduction!
  • Compressed most textures
  • Used quantization for model attributes
  • Converted models from .gltf to .glb

Referencing issues Fixes #102