I agree. And the long length of the line increases the likelihood and magnitude of any GPR/EPR and from that comes damage from ‘slow cooking’, ie modest currents rather than huge ones, but enough to cook network interfaces for example. I would say this is more likely because one end of the line can more easily be right under the cloud, inside the footprint, while the far end is well outside the footprint, thus seeing the full potential difference between inside and outside.
BTOR engineer arrived here a short while ago; in the house testing line 4.
The line has been up and down like a yo-yo; so bad that I should have disconnected it from the bonded set in fact, if I’d had any sense, as it was probably giving a ‘negative contribution’ on the whole, with all the packet loss seen at times poisoning the reliability of the whole combined pipe.
The verdict is that the short BT-to-RJ11 cable that I was using had gone bad, presumably due to the lightning cooking. I had swapped out modems but not that cable. So it means that this is my fault.