These filters will take that into account since it never has exactly been 50/60Hz. It's just the expected mean. Maybe it's present in subsequent harmonics (e.g

These filters will take that into account since it never has exactly been 50/60Hz. It's just the expected mean. Maybe it's present in subsequent harmonics (e.g., 100/150/200Hz for 50Hz) as a systematic deviation from additive white noise. But, this would be difficult in practice to reliably isolate these signals via some form of component analysis.