The problem is with the editorial taking too much effort as it is not much interesting.
Usually the quench of solution is interesting enough. Dont read the editorial just for sake of reading it, especially if you havent read the problem statement either!
hook in problem statement (something that makes me interested in the problem, like the story in problem statements)
Its an editorial, not a fan fiction for God’s sake! What “interesting” thing can you have in an official solution of a problem? I can only think of memes but they are pretty irrelevant lol.
Repeat after me-
“EDITORIALS ARE OFFICIAL SOLUTIONS TO PROBLEMS, FOR THE PEOPLE WHO TRIED , BUT FAILED TO SOLVE THE PROBLEM. THEIR PRIMARY PURPOSE, IN NO WAY, IS TO ADVERTISE AND INCREASE POPULARITY/INTERESTINGNESS OF A PROBLEM.”
You read editorial because the problem interested you in first place. Vice versa is not applicable, and it is also not the way to follow. End of Discussion.
a simple solution : taking small number examples to explain the thinking process that was gone through to come up with the solution, a gradual increase to bigger values and later a mathematical solid proof
While ideal, its not always possible.
Try writing a sample editorial of a simple, and a medium problem. I am pretty interested in how you feed the obvious to the beginners and how you break down complex stuff for beginners without making the editorial long and boring. Some things are better understood when trying ourselves. As usual, I always say, if you feel you are up for the task why not apply for editorialist and show how its done? 
drawings representing the solution to aid visual learners.
Sometimes, usually for geometry problems. You cannot have drawings for each problem, like, what will you draw? Lol. Visual learners, sorry! Please learn how to comfortable learn from reading as well 
I believe readers understanding the editorial is the main objective, not the effort the editorialist puts in making the editorial. The effort is appreciated but it’s not the main objective.
And I already went over on why this metric of readers understanding stuff is flawed in itself. Not repeating that.
In current system, all editorials are approved by setters/contest-admin/tester to make sure they are good enough and accurate. I am strictly not entertaining anything like the criteria you mentioned simply because it varies from person to person, rating division to rating division.
A method of writing editorial liked by people in 5 star spectrum may not be liked by people of 2 star spectrum. And writing editorials to appease a particular spectrum leads to an overall worse situation because now people of lower spectrum find it more difficult to fill the difficulty gap from one spectrum to another (as editorials for higher spectrum are now written with top rated guys as audience in mind rather than newcomers.).
So no, your criteria overall leads to a pretty worse situation. Feedback Rejected.
I don’t think GFG makes you memorize anything. As far as concerned with solutions I’ve seen, I’ve just seen logical explanations. Maybe you’ve come across a problem statement on GFG I’ve not come across.
Bro, did you do GFG enough? Take this as example- Tarjan's Algorithm to find Strongly Connected Components - GeeksforGeeks and let me know how interesting and easy it is
I kind of agree with many of your points. Just addressing the one where I disagree.
Drawings? Why don’t you draw it out yourself if you have such a hard time comprehensing the solution.
Sometimes it is very difficult to put into words the exact thing, or exact test case. For such cases, a drawing by editorialist is appreciated. Usually for geometrical or graph problems.
Editorials right now are actually too godamn long and dragged out.
Which I feel is fine as long as people understand. It’d not be fine if even after reading them people do not understand. Its not like the editorialist is having a gala fun time writing long editorials either.
Proof sections should be left to the reader so at least they have to put in some effort at least, or linked to sources instead of being explained in baby - steps.
Most of the problem boils down to 2 things - What is the correct approach, and why is it correct? This why becomes a major part of the editorial, and editorialist tries to impart learning through that. Telling “what” is correct can be easily done in 1-2 paragraphs. The “why” is important, and the real beauty of the editorial. Skipping the “why” means the editorial will be incomplete and a larger fraction of beginners/strugglers would not benefit from it.
The one who came close would have already known most of the proofs.
Sometimes, even after putting a lot of effort, they do not come close to actual solution one way or another. I think that will now help you understand my point of view better.
Overall, the editorials should just be shortened to the quick - explanation section. At least now some damn effort will have to be put in to understand, and if you get bored, that’s your fault. You won’t if you actually spent some time solving the problem and found it interesting.
Quick explanation section is Codechef’s solution to this dilemma. Top rated guys come, read the quick explanation and jump at relevant parts to satisfy themselves. New rated guys get a gist of the solution, make notes of crucial steps to learn and move on to main explanation. I feel current system is fine.
So, if someone is concluding that editorials are long and boring, he just isnt reading them the right way. Perhaps I can make a long problem on it someday 
)