Graded hints are useful; the last one usually is not.
Graded hints are one of the better designs in daily puzzle games, and they are also one of the easiest things to misuse. The design is simple: each level reveals a bit more, and you choose when to stop. The misuse is taking the last level.
The last hint usually gives you the answer, or something one step away from it. That converts a puzzle into a confirmation, and a confirmation teaches nothing. The useful hints are the middle ones, the ones that describe the shape of the answer without naming it.
My rule is that I can take the first two levels and no more. The first tells me whether I am even in the right region. The second narrows the route. If I cannot finish after that, I accept the loss and read the answer afterwards, which is a different activity with a different value.
This is a small piece of self-discipline, and it is worth it because the hint system is asymmetric. Taking one early hint costs a little satisfaction and saves a lot of frustration. Taking the last hint saves two minutes and costs the entire point of the exercise.
The middle levels also happen to be the most instructive. They force you to think in terms of structure rather than in terms of words, and that is the skill that transfers from board to board. The answer, by contrast, is specific to one day.
So read the hints, but set a stopping point before you start. Two levels is enough for almost every board, and the boards you fail after two levels are the ones you will remember.
The graded design only works if you respect the grade you stop at.
When a board fights back, read the hints before you guess again. The first level describes the shape of a route rather than the route itself.