A simple and easy method of detector calibration would sure be nice. I wonder if cosmic rays might be part of the answer. They are ubiquitous and uniform day and night, throughout our solar system, as far as I know. Probabilistically, short term they vary greatly but over a longer time span an average should converge to a small stable number. I was thinking a few hours count with no moderator and a few hours count with a water jacket moderator several inches thick (plastic bucket full of water) and Bob's your uncle.
Messing up the traceable
Curious what the old metrologists in the group think.