It would not matter if there is a representative sample of BW and FW divisions. They have to average fights out over the divisions regardless. There will never be anywhere near equal number of fights per divisions, so the fact these divisions were not around would only matter if the sample size was too small for a study.
In other words if there was 100 flyweight fights since 2007 and 300 MW fights, the average should still be a fairly good representation of trends.
mad props: Toxic