OK. I see what you are saying I think, which is that the stationmaster can easily rationalise why the reflection from the rear of the train arrives before that from the end of the platform. So no weird concepts are needed. That's fine, but you are only considering one observer in one frame of reference, so that is what you should expect.
Relativity of simultaneity only arises when comparing the perceptions of different observers in different frames of reference that are in relative motion. Your thought experiment doesn't do this.
Yes, you have captured my thoughts quite well except that I am not considering as a reflection but rather as an event consisting of a sudden illumination of the rear of the train. The front also would work.
The stationmaster can easily rationalize it and this is very observable to the extent that it could be set up (though not by me) as an experiment.
I think that seeing the image generated by the flash is different from a reflected beam which is how SR treats it. The difference is which frame of reference is considered for the light arriving at the stationmaster. He sees it at speed = c in his frame but imagines it as coming towards him (c + v) when he thinks about the train.
In terms of the perception of different observers both see light at speed = c, both see the same length, and both agree on speed = v.
Both agree on observable events and can calculate, based on the location of events, that the time differences are equivalent to the distance from them.
So why the difference between an analysis of the observable events and the theory of SR?