For reference, the underlying issue was that there was some stray md
metadata left on one of the devices (which I think was meaning md was
incorrectly trying to build the wrong array); deleting it with
--zero-superblock let FAI continue successfully.
While spotting that might be a bit more than you could expect from
setup-storage, I think it's error handling could be a bit more robust
(which is presumably why it said "please open a bug report" :-) )
The Wellcome Sanger Institute is operated by Genome Research
Limited, a charity registered in England with number 1021457 and a
company registered in England with number 2742969, whose registered
office is 215 Euston Road, London, NW1 2BE.