I wanted to share a resource/tool that was put together regarding human genome reference assemblies. This interactive map showcases global initiatives underway to generate reference quality human genome assemblies for diverse populations using long read (PacBio) data. This is interesting because of the 65 most recent human assemblies accessioned in NCBI, 54% use PacBio data. If you look at just those most contiguous assemblies with contig N50 > 1 Mb, 94% of the assemblies use PacBio data. It is pretty clear you need long reads to make high quality assemblies.
This is useful because it includes links to the publications, NCBI assembly accession numbers, and in many cases locations of the raw data you could use/manipulate as you see fit.