[allmost]

The concept is decades old. Casimir et al were exploring this theoretically in the first half of last century. Pulling real world usage out of random quantum fluctuations can no doubt be done, the random nature of the wavelengths/position/time/etc make it extremely difficult to convert to classical 'work'.

[LibWhacker]

No, there's nothing new about the quantum stew, but this is new work, just written up in Physical Review Letters describing a way to separate out matter and antimatter from the vacuum.