Some Enlightenment is required
As somebody who had the privilege to work on the software for, and install, the IECC signalling system for British Rail in the late 80's I would like to enlighten you on the subject of signalling. I'm sure somebody with a fine collection of anoracks and notebooks will correct me but the gist is important.
On the UK rail network, signals turn red automatically when a train goes past. There is no manual intervention. Further, the interlock system that turns the signal red will prevent an operator from routing a train into an occupied section of track, it simply won't allow the route to be set. At best it will allow the operator to specify where a train is to go, but it will not permit two trains to occupy the same block of track at the same time.
It would appear that the first train got some way into the tunnel before failing - past several signals. The other trains would simply follow on according to timetable because they had a green signal - the systems are largely automated (I still remember with fondness the train from London at 13:15 every Monday that would crash the automatic routing software in York). So the trains simply backed up in the tunnel, one per block of track, and subsequently failed as described when the snow melted.
It isn't all that unusual that the other trains were allowed into the tunnel, after all, the signallers would not know that the problem hitting the first train would also hit the subsequent ones - after all, they didn't know what the problem actually was. There was no reason to believe that all the trains would break down, and there was no rail safety issue in letting them move on into the tunnel. Especially as the alternative was to block up track in, or on the French side of, the Calais terminus. As far as they were concerned, one train broke down and as soon as it was towed out, the rest would continue on their journey.
And to be picky on a point of grammar: "The combined temperature, snow type, and humidity in northern Feance and the tunnel were exceptional - that means untested in system testing terms". To be precise, the systems were untested, not the weather conditions.
(Hello to all my ex-IECC project colleagues, with special mentions to Alistair M, GT 'taxi from Bristol' Smith and Graham Stacey).