Somewhat separately, there's the accretion disk, which again is a disk not a sphere, much like other orbiting systems like solar systems or galaxies−the gravity between bodies orbiting the same central gravity source causes them to arrange roughly into a plane, rather than all having their own unrelated orbits. We're not seeing the accretion disk directly though, but rather the light from it, and from other sources, that is able to pass around the black hole. (ie the black sphere in front of the light bulb.)
Watch that video; it explains it in a very approachable way.
1. Along with its rotation comes the fact that the black hole drags the surrounding spacetime along with it (whatever this means), including matter. So matter near such a Kerr black hole will start orbiting it automatically. Closely related(×) to this is the fact that, in the close vicinity of a black hole, you typically find a so-called accretion disk of matter that is orbiting the black hole and slowly being eaten by it, while also emitting light because the infalling matter is heating up in the process. Now, the important point is that the disk is really a disk, though(!), meaning that it doesn't completely surround the black hole in all directions, so there are (lots of) angles from which you could actually "look at" the black hole and your view would not be (entirely) blocked by the matter (and the light it emits). I hope this answers your question as to whether the light "should not […] be all around it".
2. In the case of M87 it seems like the axis of rotation is pretty much parallel to our line of sight, meaning that we're actually looking at the black hole "from above" and that our line of sight is pretty much perpendicular to the accretion disk surrounding the hole. In particular, this means we get to see the accretion disk and the black hole's "bald head" in their full glory. Moreover, since we're looking at the black hole "from above", its slight deviation from spherical symmetry doesn't matter and it still looks like a disk to us due to its rotational symmetry in the direction in which it rotates. (Think of how a cylinder looks like a disk/sphere from above.)
(×) To be precise, infalling matter often carries angular momentum (as measured with respect to the black hole's location), i.e. it doesn't fall into the black hole exactly radially but rather sideways, possibly after having orbited the black hole multiple times. This means that when it finally gets absorbed by the black hole, the latter will absorb the matter's angular momentum, too, and start spinning.(××) So the rotation of the black hole, on the one hand, and of the matter outside, on the other hand, are tightly coupled phenomena and disentangling what came first is a "chicken or egg" kind of problem.
(××) Side note: Infalling matter transferring angular momentum to a black hole is the reason why we expect most, if not all black holes in nature to carry angular momentum, i.e. to be of the (axisymmetric) Kerr type instead of the simpler (non-rotating and perfectly spherically symmetric) Schwarzschild type.
