There are occasional slip-ups when FD e.g. use the tonnage numbers from a previous trade CG for Thargoid tissue samples. On the whole, though, they get within an order of magnitude with the top tier and the result, and I think that's fairly good estimating from a limited data set.
