Brain behavior
From what I read in the brain functionality announcement ( it sounds like hashcat tracks attempts anchored around a particular hashlist.

Does that mean the hashlist needs to stay identical (name and contents) to benefit from the brain entries?
The name does not matter I believe but the contents definitely do.
I'm trying to figure out what sort of latitude I might have. eg: if I consolodate a bunch of stubborn hashes into a new file, will it know whats been done against each hash, or (I'm thinknig this) a whole new record tree gets started.
The target hashlist does have to be file-checksum-level identical, IIRC. But I'm not positive. But it also shouldn't be very hard to create a very simple test case to verify it directly.
Creating a new hash file with a different set of hashes will invalidate all previous status.