Well actually you can lower that down to a single (&sad) select SQL by using SUM as a window function as well:
select CALMONTH, INDIVIDUAL_CNT, TOTAL, PERCENTAGE, RANK, (CASE WHEN RANK<=0.8 THEN 'Fast' WHEN RANK > 0.8 and RANK <= 0.95 THEN 'Medium' ELSE 'Slow' END) as RANK_STR from (select CALMONTH, INDIVIDUAL_CNT, TOTAL, round(to_decimal(INDIVIDUAL_CNT/total)*100, 4) PERCENTAGE, percent_rank() over (order by round(to_decimal(INDIVIDUAL_CNT/total)*100,4) desc) as rank from (select CALMONTH, INDIVIDUAL_CNT, SUM(INDIVIDUAL_CNT) over () as total from ( select left(replace(FLDATE,'-',''), 6) as calmonth, count(BOOKID) as individual_cnt from SAPABAP1.SBOOK group by left(replace(FLDATE,'-',''),6))))
Performance was actually good (considering I'm using a single small table and not a complex view): total execution of 3.8s with 2.9s on JECalculate (replace and left function on the innermost select query). So if you don't have to calculate anything it should be wa(a)aay faster. It also consumed 3.6GB to execute it in my case:
ps.: Still haven't tried with WITH clause.
BRs,
Lucas de Oliveira