Homologous/Non-homologous recombination question!

Trying to work out and reason through this practice question to help me on my test.

"Briefly explain why non-homologous end joining is overwhelming favored over homologous recombination repair for fixing double strand DNA breaks in mammalian cells in G1. Why is this bias reversed during G2?"

So I reckon that non-homologous end joining is favored in G1 because although original base pairing is lost and chewed back by exonuclease before the DNA is re-ligated, the loss of base pairs isn't critical due to the imminent S phase, which will ensure the DNA is further re-synthesized, thus mitigating the effect of lost base pairs.
However, for G2 nonhomologous recomb. is favored because in this method base pairs are re-added before getting re-ligated, ensuring no bases are lost. This is critical because the next phase is mitosis, and all of the DNA must be synthesized to this point; any loss of DNA before mitosis would probably lead to mutations when the cell eventually divides.

Is my reasoning accurate? Do holiday junctions add to the explanation or does this suffice? Please chime in and discuss!