This study is done by Broadcom. 'Something is rotten in the state of Denmark' one would want to state here in response...
The study is done by Broadcom because PhyR is a proprietary Broadcom technology. It only works between a Broadcom DSLAM and a Broadcom chipset modem.
Therefore, I'd argue that these PhyR technologies are only good for a temp problem, like occasional noise.
Why only good for a temp problem? It's basically G.INP.
It runs all the time. There are very few lines that wouldn't benefit from having it. It does wonders with different types of noise
In the UK it's only deployed on some BT Wholesale ADSL DSLAM's.
Having always been on a Huawei cabinet with no ability to disable G.INP you won't know how much it's likely helping your line.
With no error protection at all my line received around 2,000 ES per day. With G.INP it received 1 ES per day and that was from a light ballast in my house coming on at 7am everyday.
Not only does it lower error numbers but it can increase sync speeds, lower latency and eliminate packet loss.
G.INP was a bit of a game changer with DSL. Many of the forums long term members remember it being rolled out and the difference it made to their lines.
Many of them on ECI cabinets saw the huge benefit only to have it taken away again.
There's a reason that the forums largest threads are discussing the deployment of G.INP.
Over 20% is quite a lot. How's it possible to achieve that much? And how can you tell that without it, your connection's speed would be 20% less?
Unlike G.INP, you can manually disable PhyR and see the difference.