In a seismology book I read long ago, I read an interesting quote on the accuracy of "time" measurement in the 1800's. It has some features similar to this scenario.
In trying to pin down the exact time an earthquake struck an out of the way location, the scientist interviewed several folks about what time it was when they felt it. He also asked how they set their watches. Several folks noted they set them based on the clocks in the local jewelers store in town.
When he investigated how the jeweler set his clocks, he said he sets them to exactly 5 pm, when the local fort sets off a cannon each day.
When he asked at the fort how they determined it was exactly 5 pm, the soldier responsible said he always checks the jewelers store in town to set his watch.
Moral of the story is that matching something that itself does not have guaranteed accuracy is not a good way to assess accuracy.