I think what you mean is that the spin axis of the black hole is not aligned with spin axis of the Milky Way? NB: That's about all you can say - the conclusion of the actual science paper (Akiyama et al. 2022, ApJL, 930, L12 is:
Our model comparisons disfavor scenarios where the black hole is viewed at high inclination ($i > 50^{\circ}$),
not that it is pointing towards us (which would be $i=0$ and highly improbable even in a statistical sense). The current analysis also fails to identify any preferred position angle of the rotation axis.
I have not watched the video to see whether they really claimed the spin axis points towards the Earth. I doubt it, and the published papers don't support that conclusion (though it is possible). Previous work on Sgr A* using infrared interferometry to monitor gas orbital motions within a few Schwarzschild radii of the black hole had already suggested that it had a low inclination (Gravity Collaboration 2018).
There isn't really any reason that the black hole spin should be aligned with the Milky Way angular momentum. When you look at the dynamics of things near the centre of the Milky Way they are not distributed in a uniform disk-like way. The stars that are immediately orbiting the black hole (which have now been studied for decades) have what looks to me like a random orientation in space$^1$. e.g. (from https://www.eso.org/public/images/eso1825d/).
Murchikova et al. (2019) show (reproduced below) the orientation of various other complex structures in the central $\sim 2$ pc of the Milky Way. They have a variety of orientations and none really lie in the galactic plane (perpendicular to the screen with the yellow line marking the position angle).
Indeed in a more recent paper (Murchikova et al. 2022), an accretion flow fed by multiple stellar sources is found to describe the sub-mm emission best. The final conclusion of that paper is important to the discussion here:
Thus, a key property of our Sgr A* wind-fed models (and thus by inference the Galactic Center
accretion flow itself) is that the gas does not circularize at large radii as discussed in Ressler et al.
(2020). If the parcels of plasma in the inner tens of rg do indeed have the broad distribution of
angular momenta (i.e., inclinations and eccentricities) that results from stochastic feeding by multiple
wind sources rather than the narrow distribution provided by a torus, then certain properties of the
accretion flow may never match models that consider only torus initializations.
Since the black hole at the centre may be the product of many, many accretion events with essentially random angular momentum vectors, or mergers between black holes with differing spin vectors, potentially from accreted satellite galaxies, any initial angular momentum direction could have been scrambled. In just the same way- the spins of stars in the galaxy, or the spins of binary systems, do not align with the galaxy's angular momentum vector.
It is probably worth noting that a $M_{\rm BH}\simeq 4\times 10^6 M_\odot$ black hole doesn't really have that much angular momentum. A maximally spinning black hole of this mass has $J \sim 10^{55}$ kg m$^2$ s$^{-1}$. An object of mass $m$ at the innermost stable circular orbit, just prior to accreting into the black hole has an angular momentum of $\sim \sqrt{12}GM_{\rm BH}m/c$, which is $10^{49} (m/M_\odot)$ kg m$^2$ s$^{-1}$. Thus the black hole angular momentum could be significantly changed by accreting $\sim 10$% of its current mass.
$^1$ In fact Ali et al. (2020) have shown that the galactic centre stars currently orbiting the black hole tend to fall into two groups. Both have close-to-edge-on orbits, but their position angles are at right angles to each other and are at 45 degrees to the galactic plane. i.e. Their orbital planes are also not aligned with the Milky Way disk.
