This appears to work for me. I ran Parallel HOP and I got the answers I expected. I looked at the 'particle_mass' and 'ParticleMassMsun' arrays on an AMR and non-AMR dataset and they made sense. Any other things to test? Otherwise, I'd say it's good to go!
Awesome! I've committed it in r1572. I think this leaves only a couple things before release -- a new build of the docs, me documenting the software ray caster ... I think that might be it.