[Sephiroth]

From a technical perspective, is your upstream stable at 51 dBmv or does it fluctuate - especially towards 55?

You see, it's unusual for there to be high downstream power and high upstream power. One can account for high downstream power - you're close to the cabinet or the amplifier is over-amplifying (or indeed both but I just think it's one of them).

But then I'd expect the upstream power to be lower because there's less attenuation between your CM and the first active node (the large street box with a 240V sticker). But your upstream power is high indicating attenuation somewhere.

We didn't see the full stats nor the event log. So we don't know if there's upstream noise affecting you.

Turning to the Speedtest result you've shown. Blank. Is that right?

With regard to the AE - the Modem Mode in the SH is the only way to run that wretched thing, IMO. Only then can you be sure that no other functions are slugging your connection. If it ran even slower in Modem Mode, that was just an unrelated circumstance, due either to poor upstream (we'll check that when you post full stats & event log) or downstream difficulties due to high power received at the SH.

In summary, I believe you have more than one problem (subject to sight of full stats and an answer to the upstream power question). The upstream power, if stable, is not out of spec. But if the upstream attenuation is high, then I wouldn't like to think what your downstream power would be if it was attenuation in your coax to the street box.

Finally, It may well be the street box amplifier that's playing up. They're making all sorts of changes at the VM end and this generally needs a rebalancing of outputs at the amplifier as lasers are renewed and so on. The visitng engineer will slap 10 dB of forward path attenuation on and it'll be worth seeing what performance difference that makes. But if you have an upstream issue, the engineer is usually unable to fix that and would have to refer matters to Networks.

Hope that helps.