OK. When you said “shown on meter”, I thought you meant TFS-100.
I meant to do this last week to help out people who noticed a change after the coefficients were corrected last week. Here is my experiment:
I injected a known 20.0Hz signal (verified by oscilloscope) into a GEN 2 unit after Franz fixed the coefficients last week. Using the published K and offsets which use the equation GPM = (freq + Offset)*K, I filled out the table for every meter in the app to verify that it converts correctly.
It would appear that they all work correctly now. For your situation specifically, 18% under counting on a paddle meter is not that atypical…particularly if it is older.
Hopefully this helps.