Shout out to bonkers who just replied with the information. I'm still trying to crush the ESB data, there is a lot of it in the spreadsheets, however ignoring the highlighted inconsistencies around sample size, sample target, and age of the original profile, it does appear the original 52/39/9 split came from real measured data.
CRU publication on smart meters. Covers multiple responses from suppliers at the time, especially around the load profile. Some really good information in there and I wish the CRU themselves had provided this when asked.
ESB load profiles. Lots and lots of spreadsheets and code numbers. Taking a random entry I get a split of 64/22/13, but with so many numbers, that is kind of meaningless.
CRU publication on smart meters. Covers multiple responses from suppliers at the time, especially around the load profile. Some really good information in there and I wish the CRU themselves had provided this when asked.
ESB load profiles. Lots and lots of spreadsheets and code numbers. Taking a random entry I get a split of 64/22/13, but with so many numbers, that is kind of meaningless.