I think it is really hard to unpick all the different factors.
Schools progressively closed from early July (private schools first), but, because of isolations, did so in a piecemeal rather than 'big bang' way. So a school with a published end of term date on, say, the 23rd, may have in fact been closed to the majority of pupils a week earlier, and to a significant minority for several weeks.
Football will have had an effect (noticeable that Wales has had a rather different pattern of cases).
Progressive openings may have had an effect, but again this may have been mitigated by widespread isolations and caution given the rapidly rising case rate (so a fall now may in fact by itself trigger a further rise, as more people feel confident to go out to larger events).
Weather has an impact on the choice between indoor and outdoor activities.
Testing is likely to decline, for 2 reasons - schools no longer asking for 2 LFTs a week, and also not picking up symptomatic cases and asking for tests (yes, we have not always been successful in asking parents to have their children tested, but some at least have been prompted to test by a child being sent home); and also the wish not to find out about a positive before a UK holiday.
The difficulty in understanding it also makes it difficult to enjoy - the lurking feeling that some mistake may be found, and also its very unlikeliness meaning that we cannot predict when it will reverse.