A related post was What is polarisation, spin, helicity, chirality and parity?
In $m_\mu<<E$ region, $m_\mu$ could be treated as massless particles and the conservation of helicity indicated the polarized cross section dependent on $\theta$.
In $m_\mu\approx E$ region, $m_\mu$ could not be ignored but that the conservation of spin made the polarized cross section independent on $\theta$.
However, notice that in both case, $m_e<<E$ the electron was in the relativistic region, and it was a bit hard to see why in one case the computation used helicity and in another case the computation used spin, where the treatment of the electron's amplitude was not explained.
Why not use the spin conservation for $m_\mu<<E$ and helicity conservation for $m_\mu\approx E$? What happened to the helicity and spin conservation in the $e^+e^-\rightarrow \mu^+ \mu^-$?