thats a good point: the fact that testing on a volunteer soldier and on army and CIA researchers, seems to suggest they wanted to compare effectiveness with the general population, since random members from the population might not be psychologically representative for foreign soldiers, secret clearance'd scientists and spies. So from this perspective testing on people of similar function / caliber from your own accessible population might seem "logical" if unethical...
Something tells me keeping statistical integrity wasn't a major factor in testing on these soldiers. i think soldiers just tend to be better test subjects for these purposes as they are more obedient, willing to sacrifice their lives for a bigger cause, take instruction without asking too many questions, keep secrets and no doubt they have more mental fortitude and a higher pain threshold to withstand multiple rounds of various chemical attacks.
I am pretty sure that facility where he witnessed a sarin attack on a soldier was a UK facility.
"Microbiological Research Establishment at Porton Down in Wiltshire"