The paper isn't particularly mass-dependent. it contemplates AGN masses down to the 10 M_sun stellar black hole in a previous paper by the second author and ot

The paper isn't particularly mass-dependent. it contemplates AGN masses down to the 10 M_sun stellar black hole in a previous paper by the second author and others. In general stars, stellar black holes, and bigger black holes have random spin axes. Central black hole binaries exist but their spins can point in random directions compared to the plane of the binary orbit; and the plane of the binary orbit is not necessarily aligned with that of the host galaxy (if discoid). As you say, galaxies can merge and in ellipticals the dominant rotation curve may be radial (blobs of (possibly star-forming!) hydrogen gas and the like cycling in and out rather than around some axis). Black hole-star binaries exist (with stellar multiplicities up to perhaps five) too, and rotation is faster with greater multiplicities, but the spin axes need not be aligned or anti-aligned, and the several stars need not all orbit in the same plane. This is an active area of research in galactic evolution. Consequently, the spilling of gas from a star to a BH companion, especially in a triple+ (and especially considering intrinsic-variable stars' structured outpourings), could lead to an accretion disc that is highly non-equatorial and/or retrograde. The paper press-released at the top contemplates the forcing the innermost part of such discs onto a more equatorial and prograde orbit, and how that warps, tears or ablates the disc compared to a "gentle" picture of a uniform disc in prograde equatorial orbit around a slow-spinning BH. > does frame dragging cause misalignment No, Lense-Thirring torques "cure" (parts of) a non-equatorial accretion disc by pulling inner elements of it into the equatorial plane. Going a little further than above, a non-equatorial disc could arise when an eruptive-variable companion spits out matter that sails just above or below the pole of its companion BH, even if the spin axes of both were completely parallel. Probably even more likely is that a large eruption spits out lots of matter and the near region of its companion BH only subtends a fraction of it. That might lead to a complex accretion structure that might settle into multiple discs. Lots of observatories are looking for that kind of thing. :-) ETA: Bardeen & Petterson (1975), The Lense-Thirring Effect and Accretion Disks Around Kerr Black Holes ends with: "We remark that both Shakura (1972) and Katz (1973) have made the suggestion that tilted disks may form in binary systems if the optical star rotates around an axis which is not aligned with the system’s orbital angular momentum. Wilson (1974) has pointed out that an asymmetric supernova explosion may change the direction of the orbital angular momentum." This also gives you an idea of how old this idea of LT precession "untilting" non-equatorial accretion discs is (I believe it originated with Bardeen & Petterson). The broad question, even older, is "why are accretion structures around stars and Saturn essentially always equatorial? Does that apply to black holes too?"