Understand variance of sampled Gaussian mocks

Gaussian mocks seem to be noiser than e.g. lognormal mocks on large scales when sampled with readout:
https://github.com/adematti/mockfactory/blob/7df2a96bee83db561a64188a2c5491777852c2ce/mockfactory/tests/test_gaussian_mock.py#L118
even with unitary amplitude.
Most of the scatter disappears when removing RSD. To be investigated.