> Last year lots of people complained about a simultaneaous reduction of ANC across several brands and models brought with firmware updates.

Was this ever verified quantitatively? It sounds like the kind of thing that could easily be generated by group confirmation bias.