From what I have seen of the spreadsheet that gives speed estimate uplifts, each cabinet will give the same uplift in speed to every connected line.
As far as I can see, the speed you get from ADSL is related to the length of the line between your premises and the exchange, via the cabinet.
The speed you get from fttc is related to the length of the line from your premises to the cabinet, regardless of how far the exchange is from the cabinet.
So how can it possibly be that all lines connected will get anything like the same improvement? You might be two miles from the exchange and 50 metres from the cabinet - that would give a vast increase. You might be two miles from the exchange and one mile from the cabinet, which ought to give a smaller improvement.
The speed estimates ought to be by postcode, calculated on the distance from the cabinet, not the same estimate for every line attached to a cabinet.
The problem is that the logic of the speed estimates seems nonsense, but it is used as the basis for more than just estimates. They make services available or not based on these flawed estimates.