Because then they can’t conclude that MBP ever does cause herpes either.
All they have is a bunch of babies who have herpes and all had MBP. To make any conclusion, you need to compare that number to the babies who did not have MBP and say there is a higher incidence, and then account for other possible differences in the samples, like a likelihood that the MBP families also have other small kids in the family who would expose them, or other possible differences.
Until you do that, it is a literature paper, not a science paper.