Why is NeRF (or NeRF-based techniques) not on the leaderboard?

[jlartois]

Can anyone shed some light on why certain techniques (like those inspired by MVSNet) appear so abundantly on the leaderboard, but others, like MSI (multi-sphere images) or NeRF (neural radiance fields) don't?

I think that the depth maps generated by for example instant-ngp or MipNerf are of high quality and could result in a top position on the leaderboard. (I assume Colmap is used for the SfM). So, the leaderboard feels a bit incomplete.

Has anyone tried to apply NeRF on this dataset and compared it to the methods on the leaderboard?

I appreciate any insight on this.