Computation speed of neutron HP

Dear G4 team,
I am curious about the efficiency of the particle HP model and I would like to know if someone has done comparisons with MCNP for neutron transport. I know that the Doppler on the fly certainly has a huge impact on the computation time since cross-sections are sampled at each collision. Hence, I guess G4 and MCNP could maybe be compared activating the SetNeglectDoppler variable (removing on the fly Doppler).
I am curious about two questions:

  1. By turning off the Doppler on the fly, am I really turning off Doppler Broadening or only the sampling of the cross-sections at each collision ? Maybe there is a process that Broaden cross-sections at initialization of the calculation when SetNeglectDoppler is activated?
  2. What is the efficiency of G4 calculation speed for HP neutrons compared to MCNP for example? Is it similar?

Thanks a lot for your answers,