ie if you have a solar panel on your roof and mostly use it to reduce the power you take from the grid, all you will see in the national numbers is lower demand, not increased production.
I think the estimate is that maybe as much as 5GW of UK wind nameplate capacity is unmetered, so it's a minority of wind power, but it's a big minority.
There's no central register of unmetered power generation, the assumption used is that if on a windy day mysteriously we're buying 25GW from the grid across the country, while on a similar but calm day it was 30GW that suggests 5GW of electricity is not being purchased when it's windy, implying 5GW of unmetered electricity supply from wind.
And there are presumably other embedded systems making up some of that 5GW estimate. Farms just seem obvious because 1) I've seen farms with wind turbines 2) They're industrial scale electricity users, often occupying terrain very suitable for a wind turbine
( The suns out! Put on the washing machine/dish washer etc ).
Bottom line - not all the local solar production is measured at the grid level - much of it results in demand reduction which can't be directly attributed.
I think this will only increase - either due to local batteries, or more electricity driven heating systems ( heat pumps for example ).
And this isn't just homes - if you have a farm and have a big water reservoir which you fill by pumping water - and you had some solar panels - you'd just run the pumps when it was sunny - there is no particular need to run them at particular times ( same if you had a wind driven pump ).
We can forecast the embedded generation in real-time, but it’s only post event once the meter data has been accounted for that we know what actually happened.