a little more details on the paper provaxx refers:
for 18-49 and 50-64 groups, risk of hospitalization decreases by 10-15 both in april-june and june-july time intervals, and death rates by 16x (with 18-49 in april-june being 30x)
for 65+, hospitalization decrease in april-june is 13 and in june-july 8, death decrease 16 and 10.
So, there is substantial decrease for 65+ and in one measure for 18-49 due to a combination of delta and time. It is possible that part of the decrease is due to an increasing number of previously infected among remaining unvaxed and more active people moving from unvaxed into vaxed column.