Progress on the connection between spectral embedding and network models used by the probability, statistics and machine-learning communities