The default argument xyz_dtype of the function read_las in io/las.py which is called by PyntCloud.from_file is float32 .
However, it may cause the difference in precision, because the type of data is float64 (double) in matlab if you use lasFileReader.
In order to eliminate the difference, you can write it like this.
cloud = PyntCloud.from_file(pointcloud_path,xyz_dtype="float64")
I think this issue needs to be mentioned in the docs.