Staple jobs in our local section workshop at RARDE / DERA / DRA was producing metalwork components for optical and IR research equipment including (lots of) one off and special purpose lenses. Normal practice was to use appropriately modified versions of the flat bottomed thread form whether Whitworth, UN or Metric depending on what existing equipment the lens in question was to mate with or what was appropriate to the rest of the design.
I'd be unsurprised to discover that, even when I left in 2004, most of our in house components were still basically designed to BS 1618 : 1949 and its successor standards.
Clearances on threads used with optical components tend to be different, usually larger and assymetric, to the official standard tables to reduce the chance of jamming and allow a tiny bit of float for things like lens mounting rings to settle correctly against the curved lens surface. Making things dead nuts to small diameter standards puts impossible symmetry requirements on the optical surfaces if the rings are to seat properly. Odds are such rings will come loose in service.
I suspect SRB used book form threads as their filter carriers were much more prone to jamming than the ones we made in house.
You haven't lived until you've had maybe £150,000 in mid 1980's money worth of one off IR telescope jam up partway down the Wreathall mount on a TICM 2 thermal imager. 16" long, 11" diameter front element, and several germanium lenses inside = heavy, serious heavy. Cuddle lovingly, not hold in hands.
Clive