What's an isentropic fan-shock?
The first bit of the moveable front ramp was carefully shaped to give a sequence of weak shocks that reduced the Mach Number so gradually that shock losses were minimised. This was close to an isentropic process, hence the name. The geometry was arranged so that as the progressive shocks were generated and the Mach angles and shock angles changed the weak shocks tended to 'focus' on a point just ahead of the lower lip. This then became effectively a single 'shock' at that point. Hence isentropic fan shock.
So the lower lip forms a normal shock and the airflow goes subsonic immediately behind it, the supersonic flow above somehow collide and form a shock between the ramps? I understand how the subsonic and supersonic flow coming together would reduce the average velocity -- I'm still surprised the gap between the forward and rear ramps wouldn't act like a divergent surface and cause the supersonic flow to accelerate rather than come down to subsonic speed.
The shock from the lower lip would, on its own, give subsonic flow across the intake, but the change in flow direction where the flow off the solid ramp started to traverse the gap (where Dude's drawing shows the flow going into the void) produced an expansion 'fan' that accelerated the flow in its vicinity and this gave supersonic flow in the upper half of the duct but there was a shear across the height of the duct there. The total
effective duct area however was convergent back to about the leading edge of the rear ramp, so the Mach Number reduced continually up to that point. Then the 'terminal shock' brought the flow down to below Mach 1 and from there on the divergent subsonic duct did the usual deceleration job. The whole point of the intake geometry was that the purely aerodynamic boundary between main duct and ramp void was infinitely flexible in shape, which made the design very tolerant of flow disturbances.