Earlier quoted context omitted.
> But my few MS experiences were not markedly better than my best other experiences. This is just as you would expect, given sufficient samples, from any classification process that really doesn't want false positives but isn't too concerned about false negatives.
Perfectly on point. I feel the question that might have missed in this interview is what kind of coverage do they manage yearly. And by region. I expect France should comme much closer to 100% than say, any south Asian country.
If they were uniformly sampling this would be a disaster, but presumably they have a (probably opaque) system for funnelling down to a manageable number.