This is the only reason I don't like long contests, if we get stuck on some problem, we can't stop thinking about it for the rest of the contest period. We could have utilized this much of time in solving other problems and learning new concepts. Also, most of the time, after reading editorial, we get to know the problem was not worth the time we wasted in it (or because WA was due to some silly mistake). Also, I feel long challenges are bit unrealistic because in interviews and other coding competitions such as ICPC, codejam etc, we don't get this much amount of time.

In my university, seniors and top coders arrange topic based private long contest in vjudge for practice of juniors, mostly by combining problems from different OJs with a large difficulty range. Usually they have 20+ problems and 7+ days. Those contests have been extremely helpful for me. Long contests allow you to think for much longer time per problem and instead of going to editorial directly, you can give up on a problem for now, go back to that problem after trying some other problem(s) and try with a fresh concept / idea / approach, all while competing with your peers / classmates. I am not sure if spending too much time behind a problem is good for improving CF rating, but I am sure it is extremely satisfying to solve a much harder problem in multiple tries.

I really like the format of long contests. Some people seem to think that if you spend a lot of time on a problem it's somehow not worth it, but I say that if you were able to solve the problem quickly it was too easy for you. When I just started competitive programming, these contests helped me a lot. Often I gave up on a problem too easily, but the long format of long contests showed me I don't need to do that, and I can't expect to come up with everything in 30 minutes.

My only problem with CodeChef Long is shaky problem quality, and that's why I only occasionally participate.

i would suggest start solving in last 3-4 days.

My opinion on this is based on my goal. I want to get better at short contests, because that is the format of every somewhat prestigious contest I am eligible for.

Doing long contests to improve my short contest skills didn't turn out to be efficient for me. I wasted 10 days thinking about problems and then next 10 days reading editorials to find out the solution uses some random data structure or technique which you won't find often in short contests.

In the end I would say, it depends on what you want to get better at. If you want to get better at solving problems within typical 2-3 hours duration, doing longs to improve on that is extremely inefficient way to do it.

Yes Long Contests are really helpful, Long contests will give you ample time to think about "how to solve a problem in various ways " and allow you to learn new concepts and apply within the contest, thus directly or indirectly improving your problem-solving skills

When I started with competitive programming, I only gave codechef Long Contests and gave like much of my time to solve those problems. It has definitely increased my problem-solving skills and knowledge but now I believe that spending time in a short contest is much more useful and realistic as all the contests are held in a short format. You don't get the required speed and accuracy through Long Contests. You can learn many important most frequently used concepts through codeforces practice section and short contest rather than giving your 10 days on questions which you would rarely use. But Long Contests are good in developing determination to solve a problem. If you are a beginner I would recommend you to short contests only and when you have gained some knowledge you can invest some of your time and develop thinking skills. The most important thing in Long Contests is that you think about all the ways to deal with the problem and that can help you in the future.