This is slightly disingenuous.
"Might Work" - 99.999% Chance
"Might Not" - 0.001% Chance
For two identical memory modules purchased from Newegg in the same order not to work together, you would have to get so, so unlucky. They pick the modules out from the same box, likely they would come from the same production line and would have sat next to one another on the same factory conveyor belt.
You would have to be buying the very last module in a pallet that had been sitting there at Newegg for 2 years, and to get your other module they open a brand new pallet that just arrived today. Even then I'd say they would still probably work totally fine in dual channel mode.
Silicon always has variation. From a silicon wafer, as I understand it, the middle parts are sold to enterprise because those are the good parts. Consumers get the edge parts, the parts that aren't broken but also not the greatest. I would therefor also expect the edge parts to have a bigger variation between them. Either way, no matter what part of the wafer a chip is made of, it has variation, compared to every other chip. Does Newegg know that every RAM stick in the same pallet came from the same wafer, for example? I bet they don't.
Even in the olden days, 20 years ago, certain CPUs made on a certain week clocked higher than anything else of that model. And that chip quality wasn't repeated.
You can clearly see the difference in silicon variance with CPUs. Same model and everything but might require 0.05-0.1 volts less for the same clocks. These are generally the 1-10% golden samples.
I think you are disingenious since you discount silicon variance. Among other things.