I'm sorry I've not read all 7 pages (to date) of this - just read the original story, the RAIB report and scanned some of your comments, so don't shoot me down if I repeat something that's already been covered.
My first comment is that I'm particularly concerned by the ability to move a train with doors open. Firstly because there should be (and normally are) interlocking safety measures to prevent it, and secondly because there should be a check in such a situation to make sure the train is in a condition to move before it is moved - given that it had suffered some significant techinal failure in the first place. I'm not speculating more than that until the full report comes out.
For my second point, I draw your attention to the Huntingdon-Conington incident of July 2005. Here's a bit of background: A down freight service failed with a seized traction motor just south of Huntingdon station on the fast line ni the middle of the afternoon. A 66 had been sent to assist with the situation. At the start of the evening rush hour, a down GNER service was routed on to the slow line, past the scene of the failed freight unit (a 60 for anyone that cares), and as the 91 was being routed back to the fast line, it snagged the wires. Headspan catenary - all 4 lines down and out. North of this, an up GNER service came to a halt at Conington in the same electrical section due to the loss of power. The outisde air temperature was 32 degrees C. I know - I was there. FCC had trains caught powerless at St Neots, Huntingdon and Abbots Ripton.
What happened next was that the operators involved in this incident (EWS, GBRF, GNER and FCC) and Network Rail had to work out how to get anything in there to assist. The lines were all blocked with failed electric trains (the NR guy I spoke to at the scene said "It's like a game of chess trying to get anything in there"), with 2 diesel locos in the mix - one of which was immovable with a seized motor. So the 66, instead of assisting the frieght train, rescued the GNER set that snagged the wires. It took a long while to sort that out. In the mean time, up at Conington, without any shade, the temperature inside the Mk 4's was so high people were feeling ill and 2 women had passed out. Some passengers took it upon themselves to break emergency windows for air. At this point, I don't blame them. However, other passengers seeing these open windows decided to literally jump out and on to the track!
The trouble was - by this point, the GBRF 66 was now at Conington, but no coupling up could take place, let alone move the set while members of the public were literally WANDERING UP AND DOWN THE ECML!!! The up line - while powerless in terms of OLE was very much still an operating railway.
This incident near Kentish Town has a few similarities - passengers being delayed in a train with no provision for ventilation when the power fails, deciding to take matters in to their own hands and in doing so, actually making a bad situation even worse.
Someone asked how it took so long to get something there to assist - I can only imagine, again, it was a similar situation to this. Simply getting something in there with failed trains blocking the way is a challenge. Even if you have the luxury of a diesel in the vicinity as was the case in Huntingdon 2005.
I'm wondering how long it will be before history repeats itself yet again, and it ends up on the evening news because a passenger, or passengers have been killed by a train because they decided to make their own way home in a situation they clearly don't understand. But I also think air conditioned vehicles MUST have a passive safety measure so that people aren't forced into such situations in the first place.