This kind of idea goes back to Baldeschi-Gelmini-Ruffini 83 and due to reinventions it goes by many other names besides "superfluid dark matter" (Berezhiani-Khoury 15), such as "Bose-Einstein-condendate dark matter" (Sin 92) or "fuzzy dark matter" (Wu-Barkana-Gruzinov 00) and several others. The latter term is the one picked up in Edward Witten et. al. 17. A recent review of the literature is in Lee 17.
It's a charming idea:
A model for dark matter made up of massive but extremely light particles, whose de Broglie wavelength is at the scale of galaxies.
The idea is that on scales above that of galaxies, the predictions of fuzzy dark matter agree with the standard cold dark matter models that work exceptionally well on cosmological scales, while on scales of the size of galaxies the quantum properties of these light particles become relevant and change their effect just so as to fix the problems that standard cold dark matter models have on these scales.
A natural candidate for such ultra-light particles are axions. (This is where Witten 17 gets in, since axions are a generic prediction of string theory.)