in problem https://mirror.codeforces.com/contest/1579/problem/D

my previous solution has core implementation https://mirror.codeforces.com/contest/1579/submission/338348122 in this i was taking two biggest numbers a,b and saving them min(a,b)*times in answer vector but it gave wrong answer on 2nd test case

this gets accepted https://mirror.codeforces.com/contest/1579/submission/338349269

but when i checked the other solution they were simulating one by one pairs i didn't get what difference will it create when taking all simultenously or taking one by one why take one by one here

please help!!!