with s as
(select 1 proj_id, trunc(sysdate, 'month') dt, 3000 amount, 10 month_cnt from dual union all
 select 2 proj_id, trunc(sysdate, 'month') dt, 2999 amount, 10 month_cnt from dual
)
select proj_id, add_months(dt, column_value -1) col,
decode
    (column_value,
     month_cnt, amount - sum(trunc(amount / month_cnt)) over (partition by proj_id order by column_value rows between unbounded preceding and 1 preceding),
     trunc(amount / month_cnt)
    ) sm
from s s1,
table(select collect(level) from dual connect by level <= s1.month_cnt);

   PROJ_ID COL                         SM
---------- ------------------- ----------
         1 2017-12-01 00:00:00        300
         1 2018-01-01 00:00:00        300
         1 2018-02-01 00:00:00        300
         1 2018-03-01 00:00:00        300
         1 2018-04-01 00:00:00        300
         1 2018-05-01 00:00:00        300
         1 2018-06-01 00:00:00        300
         1 2018-07-01 00:00:00        300
         1 2018-08-01 00:00:00        300
         1 2018-09-01 00:00:00        300
         2 2017-12-01 00:00:00        299
         2 2018-01-01 00:00:00        299
         2 2018-02-01 00:00:00        299
         2 2018-03-01 00:00:00        299
         2 2018-04-01 00:00:00        299
         2 2018-05-01 00:00:00        299
         2 2018-06-01 00:00:00        299
         2 2018-07-01 00:00:00        299
         2 2018-08-01 00:00:00        299
         2 2018-09-01 00:00:00        308

20 строк выбрано.

     month_cnt, amount - nvl(sum(trunc(amount / month_cnt)) over (partition by proj_id order by column_value rows between unbounded preceding and 1 preceding), 0),