Light that comes within 3/2 of the Schwarzschild radius (Rs hereafter) gets captured by the black hole. Of course I don't mean that it beelines in a direct radi

Light that comes within 3/2 of the Schwarzschild radius (Rs hereafter) gets captured by the black hole. Of course I don't mean that it beelines in a direct radial direction to the singularity. Of course it takes a curved path depending on its exact trajectory, but at some point on this trajectory it hits the event horizon. Nothing that hits the event horizon can ever come back out again (according to an observer at sufficient distance). This magic 3/2 Rs radius is the Innermost Bound Circular Orbit (IBCO), and is the truthfully correct scientific principle that you are grasping for. It is not a stable orbit, but it is an orbit. In a theoretical world, if a photon had the perfect trajectory to enter this orbit, it could orbit there indefinitely, but it is unstable, so in the real world it doesn't happen this way. If a photon hit outside of the IBCO, then its trajectory is affected by the pull of the black hole, but it escapes the gravity well, and continues on a new trajectory elsewhere into space. After the photon hits the event horizon, then all statements needs a big asterisk with them. Its behavior might be observer-dependent or undefined. But at least in the conventional view of general relatively equations, it doesn't need to travel directly in the radial direction to the singularity, but it has to travel within a limited cone of possibilities, none of which lead back out. It can change its angular position somewhat as it gets closer to the singularity, but there is no orbit. No light cone within the event horizon allows for increasing the distance between a point particle and the singularity. In these conventional equations, all things that fall into the event horizon _hit_ the singularity within finite time according to a distant observer. Time has some shifty meaning to things inside of the event horizon, and maybe this is all wrong. But I am sure that orbiting the singularity inside the event horizon definitely does not happen.