To be honest, I have made a mess by using far too few git commits. Will
try to tidy things up before the next paper is submitted.

EFSOI stuff is now reaching usable status for the LETKF paper. Also
found that when mirroring the gaspari cohn I had one index wrong.