How am I supposed to contemplate the possible causes with the raw data?
I don't think any of us here have the knowledge to do that even with all of your data. I would not even know what I was looking at.
Well, asking me about the details would be the most straightforward way, but apparently there are people who don't trust my judgement, so they'd be free to go for professional assessment elsewhere.
Also, you can just do general guesses, like:
1) There are differences between the tested vials during manufacture.
2) There are differences between the tested vials after the arrive to me.
3) My analyses suck.
There are some other options too, I just picked some most obvious that can be brainstormed upon.
Then we can do thought exercises, eg. I'm going to borrow claim that was already made "such differences did not occur with old method" - then I would ask:
1) Are there data to support that?
2) Could that not have been caused simply by the fact that the impurities that are now seen with the samples that cause the bulk of the difference were not detected with the previous method?
Or nitpicking my processes.
It'd be more of a logic exercise than analytical chemistry one in my eyes, but I could be mistaken. I'm a little bit autismo about these kind of things, but that's what I'd consider a productive process.
I mean, working with me and asking me questions usually works, as it's not like I bend the data to fit my narrative. The second brand owner that I tested two vials of recently literally told me he's sending again the same batch vial and yet the results came as they did.