That's a very interesting plot. I picked up this event on one of my meteor cameras and the contrast-stretched video is here. The trail is first visible above the tree on the right at 19:28:18 UTC. It was still bright twilight but ionised trail behind the spacecraft is clearly visible along with a few flashes along the way. It would be interesting to see if they correlate with your recording.

Thank you for your interest, Nick. I have checked and the Samsung NC10 netbook is setting its clock automatically using time.windows.com. Its clock also agrees with my smartphone and radio controlled clock, so the times recorded in the plot from the system clock are likely to be accurate. I have looked at your video, which I imagine was from an SSE or SE facing camera both as the expected track was from NW-SE and my antenna is at 135 degrees (SE) and 10 degrees altitude (towards the sky high over Graves), and note a prominent brightening at 19:28:36 and in this linked screen grab . That time corresponds to the transition from the gradient to the horizontal overdense reflection in my plot. A speculation might be that is when the satellite changed from a hot but metallic body to a broader ionised trail.

The camera is in Chelmsford looking SE over northern France. The event at 19:28:36 does look more than a coincidence. Before that you were getting scatter from a rapidly decelerating object with a decent radar cross section. At that point it disintegrated into many smaller parts and you then get scatter from the stationary ion train. It would be interesting to work out the geometry of this. What sense is the frequency here? Prior to 19:28:36 is it consistent with the change in path length from the transmitter bouncing off an object which is decelerating and moving away from your receiver but towards the transmitter.  

There are lots of multiple detections of this on the Fripon network so you can piece together the track. We also know the pre-entry orbit so would be able to work out the path fairly easily.