Re: It would be interesting to see the methodologiy
"Methodology"? Getting a little ahead of yourself Phil O'Sophical :)
I would be interested to see the rationale for the study, as currently the value of the table of figures is as much value as taking the average speed of traffic across a country (such as the UK) and find that in the UK the national average speed is sub 20mph and thus conclude the UK needs more motorways!
Reading the Cable.co.uk release on the report [ https://www.cable.co.uk/news/new-broadband-league-shows-uks-average-speed-is-less-than-half--700001889/ ] it seems the primary motivation was to create a global league table - ie. provide material for marketing purposes; not to actually do a meaningful piece of research. Looking at the published data [ https://docs.google.com/spreadsheets/d/1A8LDcCLY3HN5Oqys6VxB0ug8xgroDADVIA2BeAF_tSM/htmlview#gid=0 ] we can see the study has some real problems: it only tested 354,329 distinct IP addresses in the UK, with no information on how this set was chosen and whether it can be said to be truly representative or is disproportionally weighted towards those with slower speed connections.
Likewise, thinking about the real world, we know that ADSL, FTTC and gFast services (ie. those with >50M of copper) all suffer from speed degredation over distance. Additionally, as the test is of actual line speed, it doesn't take into account the fact that people will have subscribed to different services: sub 8Mpbs ADSL, upto 18Mbps ADSL, upto 38Mbps FTTC, upto 80Mbps FTTP etc.
Finally, we have to put all this into perspective and ask the question: for the typical/average UK household what is an acceptable speed. For my household, before children and when Internet was only really used for email and browsing, 1~2Mbps was sufficient. Now with a home office, teenagers, Amazon Prime, Xbox etc. 38Mbps FTTC (with a typical line speed of circa 33Mbps) comfortably satisfy's demand.
So I think an open challenge to Dan Howdle, consumer telecoms analyst at Cable.co.uk is inorder, namely put into the public domain the full rationale, methodology and data set (*) behind the report.
(*) "This is why all of the data collected by M-Lab’s global measurement platform is openly available, and all of the measurement tools hosted by M-Lab are open source." M-Labs [ https://www.measurementlab.net/about/ ]. I note at the present time M-Labs have not obviously released the data set for this piece of work to their Google Cloud Storage [ https://console.cloud.google.com/storage/browser/m-lab/ ]