Should be helicity suppressed as well, correct?
Any Pseudoscalar (spin 0) going to two spin 1/2 particles requires a spin-flip on one of the outgoing fermions to conserve angular momentum. This spin flip introduces a proportionality to the mass of the outgoing fermion, in this case a neutrino, which is very very small in relationship to the decaying particle, thus highly suppressed.
A naive estimate, I would take the predicted rate for {K,D,B}_0 -> e+e-, and multiply it by the mass of the neutrino over the mass of the electron squared.
So
[tex]
\frac{m_{\nu_e}^2}{m_{e}^2} = \left(\frac{2.2\,eV}{511\,keV}\right)^2 = 2 \times 10^{-11}[/tex]
so much smaller than the already-small branching ratios.