I would simplify it. Don’t actually read position from the meshparticle. Recreate the same data.
-
Read the orientation, scale and location from the particle that has the meshrenderer. Store that as a Particles attribute on the emitter you want to “sample positions” to.
-
Sample a static mesh that’s using the same mesh as the renderer. Apply positions.
-
Orient, scale and offset the positions. It should now match.
If you have multiple potential meshes, either use an array on the recieving emitter, or set up one emitter per mesh if that feels easier.
Edit. Yep, that worked.
Try this, then it should be easy enough to expand to multiple particles and meshes. Have fun.