But diet isn't the only data point here, pretty sure diet plus home state/region (maybe discernable from accent and/or surname) is likely enough to determine it
> pretty sure diet plus home state/region (maybe discernable from accent
You would be wrong. There is absolutely no way to tell a person's caste based on where they come from.
> surname
Surname could be an indicator but that doesn't need questions about diet.
I know people who were raised in vegetarian Hindu households in India who became beef lovers after they moved to the US. I'm skeptical that questions about diet are particularly useful for a would-be caste discriminator.
