It switched off at 0643 UTC, 29 July. It didn't fade out; it went from 10 dB over S9 to gone in milliseconds.
[Edited the following paragraph for improved clarity]
Forgetting about TDoA for a second and just thinking about trying to locate it from pure propagation observations alone, I checked a few SDRs in Asia and eastern Europe before I started the TDoA work at around 0640 UTC, just to get "the lay of the land". At that time, around two hours after sunrise in western Europe, it was not nearly as strong in western Europe as it normally is in middle of the night. The signal at receivers in eastern Europe, where it was normally quite strong during darkness, had low or no signal. Another observation was that some kiwis in Finland and Sweden had a very strong signal.
I was starting to think that it maybe it is in Kaliningrad (per tiNG above) and was focusing the TDoA around that area. While setting that up, the TX switched off.

That's the end of that for tonight.