I was running through the lidar demos in prep for the 4.4.2 release and noticed that this demo doesn't appear to be working.
This ended up being an out of order construction of the transform matrix. The fix was to swap two matrices in the construction in order to do a flip (needed for the local up-vector of a platform) before anything else. I think that this used to work because of some issues in movement that have since been resolved.
The scope of this problem and the fix should be very small, only affecting sources mounted on platforms with a spatial offset.