(In)consistent vacuum magnetic permeability mu_0

[MischaMegens2]

Since CODATA19, the vacuum permeability $\mu_0$ is no longer exactly equal to $4 \pi 10^{-7}$ H/m. This has been addressed in Magpylib by referring to Scientific Python's MU0, see Issue #731 the new MU0 problem. However, it seems there are still some $4 \pi 10^{-7}$ lingering, for instance in:
line #91 in field_BH_circle.py:
return np.vstack((Hr, np.zeros(n5), Hz)) * 795774.7154594767 # *1e7/4/np.pi
line #364 and #389 in class_BaseExcitations.py:
self._polarization = self._magnetization * (4 * np.pi * 1e-7)
self._magnetization = self._polarization / (4 * np.pi * 1e-7)
Some of the tests have it too.
It would be nice to change these, and use consistent value for the permeability.