I think they hint at the answer in the article. They mix and match stuff between the different hives during operations so it’s assumed to be infected everywhere

I think they hint at the answer in the article. They mix and match stuff between the different hives during operations so it’s assumed to be infected everywhere within a single operation plus they detect it at a time that you can’t attribute it to a single hive. Maybe the cost of testing each individual hive is too great and the cross contamination issues too likely to have hives infected but not yet observable via tests?