Well, to be honest, I don't understand the feedback chopper topologies, because according to my understanding
-they does not ease the filter requirements
The feedback loop will reduce the upmodulated signal if it has bandwidth at that frequency. Therefore the same filter is needed in the feedback loop as in the classical scheme, since the signal/chopper frequency ratio is the same.
-they does not increase the amplifier bandwidth (like in the feedforward architectures)
Due to the mentioned reasons they don't make lower chopper frequencies possible compared to the classical scheme.

So I am puzzled here. I would like to understand what are the benefits of the feedback configurations. I would love to hear your opinion about it.