Doesn't help your upstream issues any, sorry, but the change to 10.4 MHz on the 3.1 upstream channel is not wrong and isn't going to break anything, neither are the new messages: 4 and 5 are you recovering your DoCSIS 3.0 downstream channels after outage, 22, 23 and 24 the 3.1 channels. You lost them a minute and seven seconds before you saw those messages.
I'm not sure who the engineer you spoke to is but nothing 'way wrong' with the 3.1 block change, it makes a tiny difference to the available capacity, not even 2.5 Mbps, and no-one on here is going to have any idea why it was changed: there are legitimate reasons for them to change it and to change it back if they do. Your bad speeds were due to packet loss, the packet loss due to bad signal quality as the T3s and modulation changes make clear.
Fixating on 3.1 channel widths, capacity or event codes is a distraction from the actual issue: the RF quality.
Given it's the middle of May and this has been running for so long I wouldn't be especially surprised if the FTTP overlay in Bristol is live before this is fixed: those extra cabinets are probably all over the place by now.
