That statement seems totally empty and could be true even if those 12 months in 2024-25 were cooler. 10% of days in 1991-2020 were hotter than 90% of days in 1991-2020, which is on average 36.5 days in a 12 month period.

I can't see the methodology, but it would shock me if they did not take into account the local high based on the time of year.

Edit: In the methodology section it is not clear whether they used one average or average for the date.