Update Spherical Harmonics definition to agree with AES69-2022 and for improved numerical accuracy

[f-brinkmann]

@isfmiho @petibub, I think the 4*pi must be removed at the following lines in the Spherical Harmonics converter to follow the lates standard version:
https://github.com/sofacoustics/SOFAtoolbox/blob/master/SOFAtoolbox/coordinates/sph2SH.m#L38
https://github.com/sofacoustics/SOFAtoolbox/blob/master/SOFAtoolbox/coordinates/sph2SH.m#L53

In adition, someone pointed out to me that the straigt forward implmentation of factorial(l-m)/factorial(l+m) is numerically less accurate for larger values of l and m and that this can be improved by computing the expression using the Pochhammer symbol. Bit I did not look into this myself.