I won’t go into too much detail about the nature of the passenger injury, but the injury was probably the worst-case scenario in terms of its type and the timing.
Not that this is any excuse for any flaws handling the disruption, of which there were some, which could actually become quite valuable learning points (should there be any energy to do the learning - and there should be).
For example, it turned out to be a bad idea to end up blocking up Farringdon-Kentish Town northbound by terminating trains at Kentish Town, which caused fairly lengthy queues whilst platforms were cleared. In turn, this meant that it took ages to implement bidirectional working on the Northbound line through St Pancras (due to trains being in the way), which otherwise should’ve worked reasonably well.
But another root cause of that congestion was that the first train stranded behind the train with the ill passenger was sent back to Kentish Town to terminate - but the driver expected to go back through the Core, as they hadn’t been advised to terminate the train, so they ended up changing ends twice at Kentish Town as well.
It was also apparent that London Ambulance were well out of their depth when it came to having a plan to respond properly - no appropriate resources turned up to move a potentially seriously injured person from a busy railway line for a very long time indeed. Once the necessary resources did arrive, it took them a grand total of about 10 minutes to actually assist the casualty.