I agree that the level found in the wine is inconsequential, but it's worth pointing out that studies of radiation workers have found that the linear no threshold model does a pretty reasonable job of explaining non-leukemia cancer risk, at least around the levels of radiation equivalent to, say, a multi-decade career as a flight attendant. Granted, the error bars are pretty big, and you have to get to levels equivalent to a trip to Mars for it to get appreciably above background.

TLDR: It's overenthusiastic to think the literature has concluded that low-level radiation is probably good for you, but good experiments are hard.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2634664/

Dose examples: https://en.wikipedia.org/wiki/Sievert#Dose_examples