Add worse quality test data to better test nuances in calculations for quality data

We need some worse test data to get better nuances on automated quality tests in the datastore for this, currently the totals generally add up to either 0% or 100%.  