Yes the absolute risk reduction would “only” be 8% but it doesn’t change the fact that ARR is a better measure than RRR. RRR which tells you nothing without knowing how prevalent the disease is in the untreated group – in which case you can recover ARR from the combination of prevalence and RRR. After all, if you have an extremely rare disease, knowing you have a 99%, RRR still doesn’t tell you what you want to know.