Minor pedantry: 10log(6/100) = -12.2, so you could sort of get the number stated if you assume they meant sound power.

These metamerial, interference based approaches, are not going to be broadband by the standards of physical acoustics.

This could be useful for helping cancel a particular harmonic of a helmholtz resonator.

The real question, to me, is does it still work at high transitional Reynolds numbers, or under full turbulent flow?

Source: End effects during transitional and turbulent flow for Helmholtz resonators (i.e. bass reflex ports) in high output pro-sound loudspeakers has been something I've played around with in the past.