Yes, it’s a bug. The world rotation is calculated on the fly and it does not factor out the scale, as it should do. There is no builtin function in vmath to convert a quat to the euler angles, but there definitely should be. I added both issues as: DEF-2032
The objects gets larger and larger as it spins. That’s undesired, right? (I feel like “legitimate bug discovery” is a pretty big step towards truly understanding a piece of software)
The strange behaviour you encountered is not actually a bug, but a misconception of how rotations work in the engine.
We represent all rotations as quaternions in the engine, a bit different than the more broadly known “euler rotations”. Just changing the Z-value of a quaternion can result in strange results if not fully aware what it means. There are great explanations online how quaternions work so I will not go into details here.
But a quick and short answer to how to solve it in your specific case: