Hi
What are you trying to calibrate with them?
It's pretty obvious they are just run of the mil parts that have been measured on a set of bench gear. They probably are fine for some things. You *might* get to 0.1%. That is if you checked them at the same frequency right after you received the device. A year later after bouncing around and some arbitrary change in temperature, who knows? Maybe 0.5%. Change the test frequency by a couple decades (say 1 KHz to 100 KHz) and all bets are off for any of it.
Of course this all assumes their unknown gear is calibrated to better than 0.02% accuracy on the reading made (not full scale). That could be a great leap of faith.
Next layer to the onion is that one guy might do a pretty good job with them. The next guy, selling (or even building) identical items is not quite so careful with this and that. The ones I get from the first assembler are good to 0.1%. The ones you get from the guy at the next bench are at 1%.
Bob