There is also http://maps.thinkbroadband.com/#!lat=54.628258777628... where we have calculated a speed for every postcode (outside NI) on the assumption every cabinet gets a FTTC service

Off Topic I appreciate, but I'm intrigued that your map shows FTTC/Cable speeds of 24+Mbps for the two postcodes (SW9 6UN & SW9 6UW) despite all 77+ lines in the two postcodes being EO and no VM available despite being available to the rest of the street.

Fully aware of the postcode area spread and also fully aware of the limiting factors of packages that people buy too.

The general aim of the map layer is to help inform ahead of the roll-outs as to what may or may not be possible. Perfection is not possible, without enabling 27 million plus lines and measuring data every few months, so one has to work with what one can produce and hopefully the map is vastly better than saying up to 80 Mbps is available just because a cabinet is present etc.

The fall off rate is pretty steep on VDSL2 hence all the moans about why different estimates can vary so much, since being 50m or 100m out on a distance estimate can mean 10 to 20 Mbps difference in speed sometimes.

There is a school of thought that also says VDSL2 does not work at all beyond 300m from a cabinet, trying to convince them is difficult, even when I can present a speed test on a 1350m line.

A new Virgin Media layer is on the way, i.e. one I've worked out from scratch. So once that appears pester again and we can see about expanding the Virgin Media exception list

And when the minimum level of interleaving, which is not uncommon, knocks about 10Mbps off the downstream, also adding 8ms to latency.

My issue is with using it to extrapolate performance over particular length lines. Unfortunately, only Openreach will have that information insofar as there records are accurate.

What would be really nice to see (although I doubt we ever will), is a statistical analysis of the FTTC lines enabled by Openreach with estimated distance and attenuation. After all, that information is available (as it's part of the sync negotiation) and I've absolutely no doubt it's available at the DLSAM. Just how much of it is collected, I don't know, but presumably quite a lot must be to enable DLM to work.

My straight line length to the cabinet is 1400m according to google earth. For 11months I have had speeds in the region of 12mb but since a BT engineer swapped my line over it has doubled to 24mb which appears to be in line with the graph linked to by tdw42

In a perfect world yes that information would be available for third party analysis.

If that information was to be available I'd happily integrate it.

In the mean time we have pronouncements from areas on coverage and others doing extrapolations based on their area, so am working hard to work in the middle and am well area of the limits of the data.