Weighting is done, but upwards and downwards, to adjust for cells (age, gender, ethnicity, income, education, location) that are under or over sampled. It's normal in all random probability samples because they are never completely, 100% proportional.
Actually, if you look in PPM markets, young people do not listen as much, but they listen. The youngest usable demo is 18-34 and much of that demo is made up of adults with jobs, families, responsibilities and less time than a teen for music exploration. They use radio as well as paid services that are pre-programmed because they are easier to use and less time consuming.
It's really easy to explain. You recruit a new sample for each week. As weeks progress, you adjust the recruit for more or less of certain cells. You get the books back and tabulate them.
Too many stations were borrowing data. Nielsen wishes to make a profit. However, they should enforce compliance, not produce a product with missing stations.
Where is there evidence of this? Recent station listings in the online trades show them to again be listing all stations.
Let's not get partisan politics involved in radio ratings.