Doesn't seem to what I need. We are a small team of two.

Thank you @heckj I think it is clear. Basically I will develop the core of the application in SceneKit and for some specific AR features I can add RealityKit to the mix. and if performance will become a problem I can do low level Metal stuff. or would you suggest spending some time and start with Metal?

I watched all WWDC regarding RealityKit and I found an interesting one about Dynamic Mesh generation 🥳. Seems straight forward, similar to Unity. I just wonder how it works with more complex meshes, say I will have 20k-50k dynamic meshes in the scene, what optimisations options RealityKit provides? https://developer.apple.com/wwdc21/10075?time=1320

Thank you. Last questions. Is there a timeline for orthographic camera in RealityKit? When it is going to be available? Are there camera controls similar to SceneKit, orbit , pan ? If not what is the work around and when to expect? Is there a mesh flatten function to reduce amount of rendering callbacks? If not what are the plans or workarounds planned?