I think HAF is a brilliant idea to allow a much wider range of developers to code against HIVE.
However, with mass adoption, I wonder how big the storage requirements will get. I see a full hived node currently clocks in around 2TB storage, suggesting 4TB storage capacity. I wonder, with mass adoption, how quickly it'll grow to an exabyte.
Also, there was mention that HAF would also offer a slimmed down version of a full node, but I didn't see it mentioned in this fork. That would be ideal.. stats on everything except post content, and retrieve the content off the chain vs the database. They mention hafah, which kinda sounds like it, but not.. I'm a little confused there.
Regardless, I love the direction HIVE is taking on so many fronts.