I wonder if I understand your question correctly. Plant fossils preserved in low energy Silt-Clay matrix are invariably carbonized as compared to plant fossils preserved as permineralized in porous sandy lithologies. As a basic thumb rule, larger plant fragments are less common than their smaller counterparts, therefore, overcounting or undercounting does not make any sense.
Some may measure the area of each taxa and use that to determine frequency. Obviously, larger parts of fronds would be considered more frequent than say pinnules of another taxon represented by detached components.