As someone who struggles with the same issue at times, I understand that. There are sometimes where a harder problem feels easier to solve than a much easier problem (which happens especially in Div3 and Div4 rounds in my opinion). However, general rule of thumb is to not skip to problem D and spend 6 attempts before solving A.

Not to mention, how often does the approach you find for a harder problem actually work. Looking at your submission history, in round 968, you skipped from solving A and B to going straight for F, which your solution timed out multiple times. Just having an approach is not enough to solve some problems, especially since CF doesn't have subtasks. Sometimes solving a problem takes trying tons of different methods to find what sticks and what observations you can gain from trying those methods; by just skipping a problem in contest if you can't immediately figure out a possible solution, you lose the chance to try and solve that problem, rather than struggle on a question where your O(N^2) solution times out when the intended solution is O(N)